Engine Design Implications for a Blended Wing-Body Aircraft with Boundary Layer Ingestion
by
Christopher J. Hanlon
B.S. Aerospace Engineering The Georgia Institute of Technology, 2000
SUBMITTED TO THE DEPARTMENT OF AERONAUTICS AND ASTRONAUTICS IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING MASSACHUSETTS INSTITUTE AT THE OFTECHNOLOGY MASSACHUSETTS INSTITUTE OF TECHNOLOGY SE P 1 0 2003 F FEBRUARY 2003 -. LIBRARIES C 2003 Christopher J. Hanlon. All Rights Reserved
This author hereby grants MIT permission to reproduce an146 distribute publicly paper and electronic copies of this thesis documen i1 whole or in part
Signature of Author: Delagafent efAeronautics and Astronautics ->ff 72 Janyary'1", 2003
Certified by: Charles Bo e Senior Lecturer, Department of Aeronautics and Astronadtics Thesis Supervisor
Certified by: \ Zoltan S. Spakovszky C.R. Soderberg Assistant Professor of Aeronautics and Astronauti Thesis Supefiisor
Accepted by: Edward M. Greitzer H.N. Slater Professor of Aeronautics and Astronautics Chairman, Graduate Office
AEqRJ, Engine Design Implications for a Blended Wing-Body Aircraft with Boundary Layer Ingestion
by
Christopher J. Hanlon
Submitted to the Department of Aeronautics and Astronautics on January 17, 2003 in Partial Fulfillment of the Requirements for the Master of Engineering Degree in Aeronautics and Astronautics
Abstract
Boeing's Blended Wing-Body Commercial Transport (BWB) has evolved over the course of its history with a traditional pylon-pod propulsion system arrangement mounted on the aft end of the centerbody. However, this novel aircraft configuration lends itself well to a more highly integrated propulsion system. It is believed that a more integrated system with boundary layer ingestion (BLI) will promote gains in propulsive efficiency and reductions in overall system complexity, thus reducing the cost of the embedded configuration with respect to the traditional pylon-pod configuration. The closest analogy to this unconventional approach is a torpedo where the hydrodynamic efficiency of the vehicle is dramatically improved by the propeller ingesting the body boundary layer. Given the geometry of the BWB a similar improvement may be possible for this aircraft. Consequently, the goal of this project is to generate a design of a concept that would exploit this effect and then quantify the impact of boundary layer ingestion on the propulsion system design. To this end, a configuration ingesting boundary layer air from the top and bottom surfaces of the centerbody is proposed based on design drivers where the potential benefits of the torpedo effect are maximized. Within this context, a parametric cycle analysis is conducted to quantify the impact of inlet pressure recovery on the performance and design characteristics of the engines. A trade study is conducted to establish the optimum propulsive cycle selection with allowances for system weight and BLI effects. A maximum fuel burn savings of 4.2% is predicted. The inlet distortion level for the concept is quantified along with the associated compression system design implications. One additional high-pressure compressor stage and a 4% fan speed increase are required to maintain adequate surge margin. Additional factors such as engine mechanical design, noise and cost are also considered from a more qualitative standpoint. With this analysis, the design space for an embedded engine becomes developed. and subsequently the design trends from a traditional propulsion system to an embedded one utilizing BLI are generated.
2 Acknowledgements
I would first like to thank my thesis advisor, Charles Boppe, for his assistance in producing this thesis. His insight and advice was instrumental in shaping the project scope and direction. Thanks also due to Professor Zoltan Spakovszky for his help in meeting the many technical challenges associated with this project and consequently lending credibility to the analysis. This project was a very complicated endeavor and would not have been successful without the help of these two individuals. Their calm, unassuming nature and precise guidance made working with them a very pleasant and rewarding experience. I am very grateful to my employer, Pratt & Whitney, for the flexibility and support required to meet this goal. Specifically I would like to recognize my supervisor, Jerry Smutney, for his genuine commitment to employee fulfillment. He is an exemplary manger and I am fortunate to have been given the opportunity to work with him and look forward to continued relations in the future. I want to express my appreciation to Boeing for supplying the resources and support necessary to accomplish the project goals. Here I want to thank Dr. Robert Liebeck for lending his time in the evaluation of the project scope, objectives, and results. His involvement added tremendous value to the project. Certainly, the importance of friends and family cannot be overstated. In this regard I feel I have been very fortunate. For providing a welcome departure from the rigors of academia I thank you all. Kelly, thank you for your unwavering patience and good humor. You have, more than anyone else, helped me to realize what is truly important. This thesis is dedicated to my parents, David and Jennifer Hanlon, to which I am very grateful. I have been blessed with parents very dedicated and engaged in the events of their children's lives and attribute my success to them. By instilling values and ethics they made fulfilling this goal a possibility.
Dad,I will neverforget the courage andpride you demonstratedin the face of overwhelming circumstances. You left an example by which I would do well to duplicate. Thank you.
3 Table of Contents
Abstract...... 2
Table of Contents...... 4
Table of Figures ...... 6
N om enclature...... 7
1. Introduction...... 8
1.1 Background: The Blended W ing-Body Concept ...... 8 1.2 Embedded Propulsion Systems...... 9 1.3 Thesis Objectives...... 11
2. BLI Physics...... 12
2.1 Previous W ork ...... 12 2.2 Introduction...... 12 2.3 W ake Analysis of BLI Phenomena...... 13 2.3.1 Induced Drag W ake ...... 14 2.3.2 Viscous Drag W ake ...... 16 2.3.3 Propulsion System W ake ...... 17 2.3.4 BLI from a W ake Analysis Perspective...... 17 2.4 Application to BW B Propulsion System Design...... 19 2.5 Thrust-Drag Bookkeeping ...... 19
3. Concept Generation and Down-Select...... 22
3.1 Project Initiation...... 22 3.2 Configuration Generation ...... 25 3.3 Configuration Assessment & Down-select...... 30 3.4 Boeing Feedback...... 36
4. Param etric Cycle Analysis...... 37
4.1 Fundamental Propulsion Theory...... 37 4.2 Parametric Cycle Results for Turbofan Engines...... 42 4.2.1 Engine Specific Thrust and Airflow Demand...... 44 4.2.2 Fan Diameter Sizing ...... 45 4.2.3 Overall Efficiency and Specific Fuel Consumption ...... 47 4.2.4 Gas Generator Core Size Impact...... 51
4 4.3 Cycle Analysis Summary...... 52
5. Propulsive Cycle Design...... 53
5.1 Boundary Layer M odel...... 55 5.2 Engine Performance...... 56 5.3 Engine Inlet Recovery & BLI Drag Reduction Calculation...... 57 5.4 BLI W eight Reduction & Trade Factors...... 60 5.5 BLI Influence on Component Performance...... 61 5.6 FPR Trade Study Implementation Tool...... 62 5.7 Trade Study Results and Discussion...... 64
6. Compression System Design Implications ...... 68
6.1 Introduction...... 68 6.2 Quantification of Inlet Distortion Effects on Stability...... 71 6.3 HPC Design Considerations ...... 74 6.4 Fan Design Considerations ...... 77 6.5 Summary and Additional Thoughts...... 79
7. Additional Considerations ...... 79
7.1 M echanical Design...... 80 7.2 Engine Noise...... 81 7.3 Cost Implications ...... 83 7.3.1 Engine Acquisition Cost ...... 84 7.3.2 Engine Operations Cost ...... 86
8. Conclusions and Future W ork ...... 88
8.1 Summary ...... 88 8.2 Recommendations for Future W ork...... 90
References...... 91
Appendix 1: Project Timeline...... 92
Appendix 2: Propulsion System Design Timeline ...... 93
Appendix 3: Boeing Project Letter ...... 94
5 Table of Figures
Figure 1.1: Second Generation Blended Wing-Body Concept...... 8 Figure 2.1: Design Links between Engine & Aircraft Analysis ...... 13 Figure 2.2: Flight Vehicle Wake Sources ...... 14 Figure 2.3: Lift-Induced Drag...... 15 Figure 2.4: Viscous Wake Generation...... 16 Figure 2.5: Wake Loss Reduction from BLI ...... 18 Figure 2.6: Effective Inlet Definition...... 20 Figure 3.1: Systems Analysis Philosophy...... 23 Figure 3.2: Second Generation BWB Concept...... 26 Figure 3.3: Baseline Boeing Upper 'D' Inlet Schematic ...... 27 Figure 3.4: Upper & Lower 'D' Inlet Schematic...... 28 Figure 3.5: Upper 'D' Inlet & Lower 'Flush' Inlet Schematic ...... 29 Figure 3.6: Aft Fan Turbofan Schemetic ...... 30 Figure 4.1: Idealized Propulsor...... 38 Figure 4.2: Actuator Disk with Stagnation Pressure Loss ...... 40 Figure 4.3: Influence of Pressure Recovery on Downstream Velocity ...... 41 Figure 4.4: Fan-Lo-High Turbofan Schematic ...... 43 Figure 4.5: Impact of Pressure Recovery on Nozzle Exit Velocity...... 44 Figure 4.6: Airflow & Specific Thrust Delta's ...... 45 Figure 4.7: Fan Diameter Sizing...... 46 Figure 4.8: Fan Diameter Size Trends ...... 47 Figure 4.9: Non-Ideal Brayton Cycle Thermal Efficiency ...... 49 Figure 4.10: Relative Impact of Inlet Recovery on Thermal and Propulsive Efficiency . 50 Figure 4.11: Relative Impact of Inlet Recovery on SFC and Overall Efficiency ...... 51 Figure 4.12: Core Size Impact of BLI ...... 52 Figure 5.1: Factors Comprising Fuel Bum ...... 53 Figure 5.2: Fan Pressure Ratio Trade Study Methodology...... 54 Figure 5.3: Engine Inlet Recovery & BLI Drag Reduction Calculation...... 58 Figure 5.4: Sources of Engine Airflow...... 59 Figure 5.5: SFC Sensitivity to Fan Efficiency...... 62 Figure 5.6: FPR Trade Study Tool...... 63 Figure 5.7: Fuel Bum Trade Study Results ...... 65 Figure 6.1: Notional Compression System...... 68 Figure 6.2: Generic Compressor Map Representation...... 69 Figure 6.3: Velocity Diagram ...... 70 Figure 6.4: Spoiled Sector Angle Influence on Stability ...... 72 Figure 6.5: Inlet Configurations Schematic ...... 73 Figure 6.6: HPC Compressor Map with Distortion ...... 75 Figure 6.7: HPC Stability Re-Design...... 76 Figure 7.1: Turbofan Weight Summary...... 81 Figure 7.2: Sources of Engine Noise ...... 82 Figure 7.3: Cost Breakdown...... 84 Figure 7.4: Engine Cost per Pound of Thrust ...... 86
6 Nomenclature rh Mass flow rate Ap Propulsor disk area BLI Boundary layer ingestion BPR Bypass ratio BWB Blended wing-body FNT Net thrust FPR Fan pressure ratio HPC High-pressure compressor LPC Low-pressure compressor NPSS Numerical Propulsion System Simulation OPR Overall pressure ratio Pa Ambient pressure Ps Static pressure Pti Total pressure (at location i) RPR Rotor pressure ratio Rt Rotor tip radius s Solidity SM Surge margin SOAPP State-of-the-Art Performance Program Ts Compressor temperature ratio T4F Combustor exit temperature (degrees Fahrenheit) T4/To Engine temperature ratio TSFC Thrust specific fuel consumption Voo Freestream velocity Vi Velocity (at location i) Vjet Jet velocity W/A Airflow per unit area a Angle-of-attack 6 Boundary layer thickness Ah Change in enthalpy AP Change in pressure AV Change in velocity 710 Overall efficiency 11th Thermal (cycle) efficiency 71C Compressor efficiency ,It Turbine efficiency Tic BLI Compressor efficiency with BLI effects TIP Propulsive efficiency IId Diffuser (inlet) pressure recovery Xi Inlet recovery p Density a Stress y Compressor stage loading y Ratio of specific heats
7 1. Introduction
1.1 Background: The Blended Wing-Body Concept
The basis of this design thesis revolves around a novel aircraft configuration called the Blended Wing-Body. The Blended Wing-Body (BWB) [15] concept is a non-traditional aircraft design in which the wing and fuselage are blended so as to create a sleeker, more aerodynamically efficient configuration that resembles a flying wing. The aircraft is the result of work conducted by McDonnell Douglas in the early 1990's in response to a NASA proposal for an advanced, high performance transport aircraft. Figure 1.1 is a representation of the Blended Wing-Body aircraft.
Figure 1.1: Second Generation Blended Wing-Body Concept with Pod-Pylon Propulsion System
The BWB design has undergone many permutations over the years and that which serves as the baseline for this project is one of the early versions given that the most recent design is proprietary. It is a large aircraft with a maximum seating capacity of about 800 passengers with seating and cargo areas contained in the center section, called the centerbody. Because of its efficient design, the BWB could consume as much as 20% less
8 fuel than conventional jetliners today [15]. Table 1.1 summarizes the mission characteristics for a large BWB concept.
Range 7000 nautical miles Passenger Capacity 800 mixed class Average Cruise Mach Number 0.85 Cruise Altitude 35,000 ft Table 1.1: BWB mission characteristics
Currently, the proposed propulsion system for the BWB is a traditional pylon and pod installation mounted on the aft end of the upper centerbody. This location provides good inlet performance and keeps the engines well away from the ground during aircraft rotation, thereby reducing the chance of ground strike or debris ingestion [1]. While this configuration is adequate, the crux of this design thesis is to investigate the design impact on the engines for a more highly integrated configuration where the engines ingest the thick boundary layer of the BWB centerbody. Such an embedded configuration has the potential for additional performance gains stemming from drag reduction and reduced overall system complexity. However, the integrated design does not come without drawbacks, and the positive and negative attributes of an embedded propulsion system will be expanded upon in the following section.
1.2 Embedded Propulsion Systems
Highly integrated propulsion systems have many attributes, both positive and negative, which have an impact on the total (aircraft & engine) system performance and cost. In general, the more highly integrated propulsion systems require additional care during design to ensure that the numerous system-level interfaces are understood and handled appropriately. The simplest and most common example is that of a fighter aircraft. Here, the need for a compact package requires the engines to be buried within the fuselage. This arrangement imposes additional constraints on the engine such as size and inlet and exhaust nozzle performance. Overall these constraints serve to reduce the performance
9 of the isolated engine but they benefit the performance of the entire aircraft as a system [2]. A commercial transport embedded propulsion system utilizing BLI will be faced with similar trades.
In proposing an embedded, integrated propulsion system the motivation is to take advantage of the intrinsic aerodynamic and structural benefits (positive attributes) of an integrated vs. a traditional (modular) propulsion system. Aerodynamically, these include a reduction in the juncture drag due to fewer intersecting surfaces and reduced wetted area hence lower skin friction drag. A trim drag benefit may be realized because it is not necessary to compensate for the pod-pylon nose-up pitching moment. Structurally the integrated system will be lighter, requiring fewer parts and fasteners as the aircraft structure is used more efficiently to encapsulate the engine. An example would be the removal of the engine pylon which would represent both a weight and complexity savings. Overall the trend is towards a lighter, less complex system which is easier to manufacture and to assemble. Furthermore, there seems the potential to realize a significant noise reduction for an embedded engine configuration. Given the recent trend in engine design to emphasize noise reduction and the future benefit of ultra-quiet aircraft in an environment of increasingly stringent noise restrictions, this potential benefit could be extremely valuable.
Embedded propulsion systems have negative attributes that tend to complicate the design process. Foremost of these involves the constraints imposed on the engine design and performance. Embedded engines tend to have reduced inlet pressure recovery and exhaust nozzle performance due to limitations imposed by the aircraft [2]. Also, the size limitations (diameter & length) often limit the optimum thermodynamic cycle selection for a given flight profile (mission). In addition, the placement of the engines often results in significant flow distortion into the engine resulting in both engine stability issues and performance loss. Together these limitations have a significant impact on the engine subsystem performance. Furthermore, the buried engines often pose a maintenance problem as a result of reduced accessibility. This in turn requires additional maintenance time for routine repairs. Also, owing to the elevated distortion levels the embedded
10 engines may require increased maintenance to address possible high cycle fatigue and operability problems with the compression system [10]. In sum the debits of such a configuration can become very significant relative to the gains in performance, weight and complexity. The challenge therefore is to accurately quantify the benefits in relation to the costs for a traditional vs. an embedded propulsion system. Only in this way can one be sure that the choice to integrate is the best for the entire system performance. For a BWB configuration utilizing BLI, the hope is that the performance gain (reduced fuel bum) owing to the drag reduction from the torpedo effect will produce a trade in favor of a BLI configuration.
1.3 Thesis Objectives
At present, the literature does not take into account how the design space for an engine utilizing boundary layer ingestion will change in relation to one designed for a traditional installation. In essence, the assumption is that the same engine would be used for both the traditional and embedded configurations. While this could be adequate it would most likely not be the optimum solution. Given the additional design considerations for an embedded engine in conjunction with the effects of boundary layer ingestion it is foreseeable that considerable differences could arise between the two engines. These differences would stem from not only the thermodynamic (propulsive) cycle design including fan diameter and core size changes but also aerodynamic considerations to address the potential compression system stability problems.
This thesis will focus on the exploration of the design space changes for an embedded, boundary layer ingesting propulsion system with respect to a traditional (pylon-pod) configuration. To accomplish this, first an aircraft configuration utilizing boundary layer ingestion must be selected to analyze. This is done in the first portion of the thesis. With this, the implications on the propulsion system design are then generated and analyzed. Before this can be addressed, some background information is supplied on the theory and physics behind the concept of boundary layer ingestion and its impact on the aircraft's overall propulsive efficiency.
11 2. BLI Physics 2.1 Previous Work
This design thesis precipitated out of the work of several previous sources in which BLI was investigated for the purposes of improving flight vehicle performance. The theoretical basis for such a concept has been well developed by Douglass [8], Smith, [6], and Smith [7]. Through the application of first principles these works develop the fundamental analysis techniques by which the merits of BLI can be understood. In the process, the level of potential performance gain is estimated which provides a basis for comparison. The Rodriguez thesis [1] (see Chapter #2) summarizes much of this before proceeding to analyze the inlet design for the specific case of a BWB with BLI. Here the estimated performance benefit for a BWB aircraft with BLI was a 1.6% reduction in fuel bum when a portion of the upper surface boundary layer is captured. This thesis expands upon the previous knowledge and investigates the engine design implications for a BLI configuration. To the authors knowledge no such study has been performed previously. Such a study will be critical when evaluating the overall system benefits for an integrated BLI proposal. Therefore, while the basis for this study is a BWB the analysis and results are applicable to any commercial propulsion system utilizing BLI.
2.2 Introduction
The concept of utilizing boundary layer ingestion (BLI) to improve the propulsive efficiency of an aircraft is not new. In fact, papers documenting the theory date back in excess of 40 years [7]. It is not the intent to discuss the details behind BLI theory; for that the reader is urged to consult the Rodriguez dissertation [1] which presents a very thorough description of the phenomena. Instead, the intent is to highlight the key physics involved so that sufficient background is supplied to both understand the design implications and the design methodologies for incorporating BLI influences in a propulsion system design.
There are several ways to look at the impact of boundary layer ingestion on the performance of an engine/aircraft system. For this purpose a perspective will be
12 presented which lends itself quite nicely to the design of gas turbine engines, namely the Reduced PressureRecovery and Aircraft Drag approach. Essentially, the ingestion of the aircraft's boundary layer into the inlet of a jet engine represents a pressure loss to the thermodynamic cycle of the engine. This pressure loss is manifested in the momentum deficit resulting from the viscous boundary layer. The momentum deficit also represents a portion of the profile drag of the vehicle in question. Consequently, a link exists between the engine thermodynamic cycle performance and the aircraft performance. It is this link that will be exploited to quantify the benefits of boundary layer ingestion and will be discussed further in section 2.2. The interrelationship between inlet recovery, engine performance and aircraft performance is illustrated in Figure 2.1.
Figure 2.1: Design Links between Engine & Aircraft Analysis
2.3 Wake Analysis of BLI Phenomena
For an aircraft in steady, level, un-accelerated flight the supplied thrust from the engine must equal the total drag of the aircraft. The drag forces on the aircraft are manifested in a viscous wake, which represents the momentum loss due to viscosity, and an induced drag wake that is a consequence of the production of lift. These two sources of drag are
13 counteracted by the propulsion system that provides a momentum flux equal to the total vehicle wake momentum deficit. Therefore, in steady, level flight the net momentum flux to the environment is zero as summarized in Figure 2.2. Each of these three sources will be expanded upon individually.
Cont rol Volume
V.
jiet iAircraft
Viscous & Induced Drag Wake Propulsion System Wake
Figure 2.2: Flight Vehicle Wake Sources
2.3.1 Induced Drag Wake
The production of lift requires that the freestream flow be turned (rotated down) inducing a reactive force up. This in essence is the circulation theory of lift. Flow turning results in a reduction of the flow velocity in the direction of flight owing to a constant velocity magnitude, since no mechanical work is expended on the flow. This AV represents a momentum loss in the direction of flight and therefore a force parallel to but in the opposite direction of the flight (drag force) [8]. See Figure 2.3 below:
14 Freestream Flow
Induced Wing Wake Flow
II II I I II a I'
AV
Note: Owing purely to flow turning, the axial component of velocity is reduced in producing lift.
Figure 2.3: Lift-Induced Drag.
From the above figure it is evident that the flow downstream of the wing has a lower momentum (in the direction of flight) than the upstream flow which is represented as AV. Therefore, as a pure consequence of generating lift, a loss mechanism exists that has no connection to viscosity. This drag, referred to hereafter as the induced drag, is unavoidable but it can be lessened with geometric variables such as aspect ratio where increasing the span of a wing (for fixed area) results in reduced induced drag. For flight vehicles the induced drag represents a significant portion, on the order of 50%, of the total drag force.
15 2.3.2 Viscous Drag Wake
The second source of momentum loss is due to the presence of viscosity in the flowing air. This wake is considerably different than the induced drag wake in that it involves entropy generation due to viscous effects. As the freestream encounters the vehicle, the viscous shear stresses remove kinetic energy from the fluid in an adiabatic, irreversible process. Within the boundary layer a velocity gradient exists where the flow adjusts from zero velocity at the vehicle surface to the freestream conditions. Boundary layer thickness is controlled by the downstream pressure gradient and the Reynolds number, with thicker boundary layers corresponding to greater momentum losses. The net result of the boundary layer is a reduction in the fluid velocity and hence a momentum deficit (wake) downstream resulting in the aircraft's profile drag which is proportional to the area of the wake and the velocity defect (AV). Profile drag makes up the remainder of the total drag force, again on the order of 50% for most flight vehicles. Figure 2.4 illustrates the viscous wake.
Control Volme
Freestream Viscous Wake
Boundary layer thickness, S Boundary Layer AV Aircraft
Note: Not drawn to correct proportions - for illustration purposes only
Figure 2.4: Viscous Wake Generation.
16 2.3.3 Propulsion System Wake
A propulsion system operates by converting the thermal energy in the fuel to mechanical power in the freestream. In the process, the momentum (velocity) of the air is increased resulting in a reactive force (thrust) in the direction of flight. A propeller or fan can be visualized as the model of the propulsion system which performs work on the freestream by transmitting shaft horsepower to the airflow with thrust power as the resulting output. The net effect for steady, non-accelerating flight is the propulsion system generating a momentum flux that exactly equals the momentum deficit caused by the lift induced and viscous effects of the vehicle.
2.3.4 BLI from a Wake Analysis Perspective
With the momentum wake of a flight vehicle now described it is possible to understand the advantages of ingesting the aircraft boundary layer for the purpose of improving performance. Here the key observance is made that decreasing the total size of the wake left by the aircraft would imply a reduction in the power required to drive the vehicle, since thrust equals drag and power is thrust times velocity. Consequently, this would imply a reduction in the fuel bum for any given mission. As has been discussed above, the momentum wake trailed by an aircraft has both lift induced and viscous contributors. While the lift induced portion is essentially fixed, the viscous portion can be reduced through several methods, most notably streamlining. For BLI, the theory is to remove part of the viscous wake by ingesting a portion of the boundary layer with the engines. This low momentum boundary layer flow is reenergized by the propulsion system and exits to the atmosphere. In this way the ingested flow does not contribute to the wake deficit and hence the realized drag of the vehicle is reduced. Figure 2.5 illustrates this principle.
17 Modular (Traditional) Propulsion System jet Freestream, V. I I vwake
AV
Embedded Propulsion System V. jet Freestream, V,,
Vwake
AVBLI < AVTraditional
Figure 2.5: Wake Loss Reduction from BLI
18 2.4 Application to BWB Propulsion System Design
In a traditional pylon-pod engine installation like the one currently proposed for the BWB, the goal is for the propulsion system to influence the aircraft aerodynamics as minimally as possible. This implies that the engine and wing flowfields do not interact. Consequently, the engine is designed to receive nearly pristine airflow [10]. In an embedded, boundary layer ingesting configuration the engine airflow is comprised mostly of lower momentum flow with the associated stagnation pressure loss; a direct thermodynamic penalty. However, the momentum deficit captured by the engine represents a drag reduction on the aircraft. In essence, the portion of the aircraft forward of the engine face can be envisioned as the "effective inlet" (see Section 2.5) with that portion of profile drag being removed from the aircraft. In this way, the propulsion system performance is debited through a decrease in engine efficiency while at the same time the aircraft drag is reduced in proportion to the amount of boundary layer flow that is ingested. Therefore, a coupling exists between increases in drag reduction and decreases in engine performance as more of the boundary layer is consumed. The net impact on fuel bum then becomes a function of both phenomena and comprises part of the focus of this thesis. This coupling is illustrated in Figure 2.1. Here the focus is on the aircraft drag - engine performance link established when boundary layer flow is ingested.
2.5 Thrust-Drag Bookkeeping
When estimating the performance of an aero-propulsion system a primary concern is the proper accounting for thrust and drag. For an aircraft, the engine and airframe flowfields will tend to interact and affect one another, in some cases severely. The result is the engine thrust and airframe drag are not mutually exclusive and can impact one another strongly. This interference phenomenon has ramifications for the system architecture in terms of engine and inlet integration as well as performance estimation. For this thesis the process by which thrust and drag is accounted for is critical to accurately quantifying the merits of an integrated propulsion system.
19 In order to bookkeep thrust and drag for the BLI configuration an outer control volume analysis was selected. This philosophy is detailed in the Rodriguez dissertation [1] as the "reduced pressure recovery and aircraft drag approach" and the central aspects will be repeated here. The crux of the technique is to envision the portion of the BWB centerbody in front of the engine as the "effective inlet" as highlighted in Figure 2.6. This philosophy tends to be conservative and interfaces well with O-D thermodynamic propulsion analysis tools.
Effective Inlet
Figure 2.6: Effective Inlet Definition
With this viewpoint, the profile drag associated with the effective inlet is removed from the aircraft drag polar and thus the required thrust from the engines is lowered by the same amount. The propulsion system performance is impacted through a reduction in inlet recovery commensurate with the momentum deficit owing to the boundary layer ingestion from that portion of the airframe.
Installation interference effects are not considered in this project. These include the influence of the engine flowfield on the span loading of the BWB and afterbody drag due to engine nozzle installation and performance. Nacelle drag is debited to the propulsion
20 system but profile drag changes due to embedding are not considered. Additional throttle dependent drags, such as inlet spillage, are also not included at this level of analysis.
21 3. Concept Generation and Down-Select 3.1 Project Initiation
The process of commencing the design project included a design review where the intent was to receive feedback and approval for the scope of the project, namely the systems analysis of a BWB commercial aircraft with a highly integrated propulsion system utilizing BLI. Secondly, the hope was to stimulate enough interest that Boeing would become involved in the project, lending insight, advice and help to meet the stated goals. With this framework in mind a presentation was conducted on March 20, 2002 at MIT. During the talk the central objectives of the project were reviewed and the overall philosophy was presented. The program objectives included:
- Exploration of performance and cost-effectiveness gains for a highly integrated, non-traditional propulsion installation = Perform system-level trade studies to determine optimum BLI configuration " Comparison of the novel concept with the current BWB configuration
Derived from these the success goal for the project was:
Quantificationofperformance and cost for a Blended Wing-Body system (airframeand propulsion system) with an innovative propulsion integrationconcept utilizing boundary layer ingestion.
The necessary process steps for the project were identified and a timeline in which the analysis could be performed was agreed to. The steps in the analysis would include: 1. Assess Literature 2. Generate Candidate Concepts 3. Select Configuration for Analysis 4. Decompose for Engine & Aircraft Analysis 5. Perform Analysis & Generate Data 6. Merge For Overall Metric
22 A timeline indicating the start and duration of the project steps was produced. Examples of the project timelines for the overall project and the propulsion system design are contained in Appendix 1 and Appendix 2 respectively.
As conceived, the project encompasses both propulsion systems analysis and aircraft systems analysis in conjunction to produce an overall system-level metric, such as $/seat- mile. Analysis of the propulsion system will concentrate on the impact of BLI on the engine performance and design. The aircraft analysis attempts to quantify the aerodynamic impact of the highly integrated propulsion system as well as the impact on total system complexity including both manufacturability & maintenance. Pursuant to these objectives it was natural to proceed with the project along two paths, one related to propulsion analysis and another concentrating specifically on the airframe, with associated connectivity as required. This modular approach is represented in Figure 3.1:
A Cost ($/seat mile) Engine Subsystem Analysis
Cyce & Dsign
A Weight System Level A Fuel Bum Influence Coefficients
0erablity, Vibration, HCF, Noise
Aircraft Analysis A Drag A Weight A Maintenance
Figure 3.1: Systems Analysis Philosophy
23 Engine Subsystem Analysis
The analysis of the propulsion system concentrates on three fronts. The first and foremost goal is to investigate the propulsive cycle impact due to the presence of BLI. This includes cycle efficiency, TSFC and specific thrust deltas from the baseline pod-pylon configuration. This parametric data provides a solid base from which to extrapolate performance trends and illustrate the thermodynamic impact of BLI. In addition, a trade study is conducted to determine how the optimum fan pressure ratio (i.e. bypass ratio or diameter) would change when the embedded configuration boundary layer ingestion and weight reduction effects are incorporated in the analysis. From this optimum cycle an estimate of the fuel burn reduction for the BLI concept is made which represents an operations cost savings. The second focus is to investigate the aero-mechanical design impact of distortion on the compression system. This encompasses engine operability considerations and turbomachinery design (i.e. fan & high-pressure compressor). Most importantly is the need to quantify the level of total pressure distortion present and then deduce the design ramifications. The third front of the project addresses additional considerations regarding vibration and possible high cycle fatigue issues as well as the potential noise benefits of the embedded configuration. In total, the output would be trends in the attributes & performance for a BLI configuration engine design with respect to a pylon-pod configuration engine design. In this way the design space for an engine utilizing BLI is framed.
Aircraft Analysis
The analysis of the airframe would focus on the impact of the novel, highly embedded propulsion system on the aircraft system-level metrics. Here the influence of the engine flowfield on the aircraft aerodynamics would be investigated. Also, the weight and complexity reductions owing to the integrated propulsion system would be quantified. In addition, the potential maintenance cost influence stemming from the inherent accessibility issues would be explored.
24 Candidate concepts are generated and the analysis is applied. The description of that process follows.
3.2 Configuration Generation
With the project scope developed the next stage of the project involves candidate geometries that form the basis for analysis. In doing the preliminary analysis it is very clear that in order to maximize the benefit of BLI it is necessary to capture as much of the boundary layer as possible [8]. Theoretically, the most efficient system would be one where the entire wing boundary layer is removed by the engines. In essence, the entire wing (top and bottom) would be covered by an inlet capturing the entire viscous flowfield. From a complexity standpoint this configuration was deemed impractical so efforts were aimed at similar goals but with more realizable concepts. However, this notional arrangement would represent an upper bound on the performance of such a system.
Keeping with the interest of capturing large amounts of boundary layer flow, the goal here is to use both the upper and lower surface boundary layers for engine ingestion. In comparison, Boeing's current candidate BLI configuration, with an upper 'D' inlet, removes airflow from the top surface only as illustrated in Figure 3.2.
25 IGOSFT
280.FT
2MDFT 4 DR
Figure 3.2: Second Generation BWB Concept
Consequently, the performance benefit could theoretically be almost doubled owing to twice the boundary layer airflow ingested by the engines. Following from this philosophy several versions are generated for analysis. A list of the notional candidates follows below:
1. Boeing 'D' inlet with upper boundary layer removal 2. Upper and lower 'D' inlet 3. Upper 'D' inlet with lower 'flush' inlet 4. Aft-fan turbofan with upper and lower 'D' inlet
Each of the configurations listed above is now described in detail.
Configuration 1: Boeing 'D' inlet This configuration represents the current model with which Boeing is pursuing an investigation of BLI. Here it serves as a comparison against which the novel, more
26 embedded concepts can be appraised. In this concept, the engines remove boundary layer flow only from the top surface of the BWB center body through a 'D' fashion inlet. With only top surface removal, the full potential of BLI is not realized. However, this does represent one possible permutation and therefore is included in the analysis. In addition, the available data on this configuration serves as a convenient framework from which the study could be based. Figure 3.3 is a representation of the Boeing configuration.
Propulsor Upper BL removal BWB Airframe
Figure 3.3: Baseline Boeing Upper 'D' Inlet Schematic
Configuration 2: Upper & Lower 'D' Inlet Here the removal of the upper & lower surface boundary layers is through two 'D' type inlets on the upper and lower surfaces. This configuration represents a direct extension of the Baseline Boeing 'D' arrangement with the emphasis on increasing the profile drag reduction. Use of two 'D' inlets presents some operational challenges however. Foremost is the increased risk of ground contact with high angles-of-attack during approach and landing. To alleviate this problem the landing gear arrangement may have to be modified. Also, with the lower 'D' inlet having such close proximity to the main landing gear, there
27 could be a tendency to ingest debris, water, and birds from the runway representing an operations risk. Figure 3.4 is a representation of the upper and lower 'D' configuration.
BWB Airframe
Propulsor
Upper BL removal
Lower BL removal
Figure 3.4: Upper & Lower 'D' Inlet Schematic
Configuration 3: Upper 'D' Inlet and Lower 'Flush' Inlet Here the boundary layer removal is from both the upper and lower surfaces of the center body. The upper removal is through a 'D' inlet while lower removal is facilitated with the use of a 'flush' inlet. The flush inlet has no vertical protrusion from the bottom of the fuselage and as such causes no rotation problem (i.e. ground contact) for the aircraft during takeoff and landing. Also, it is believed the flush inlet will provide a lower risk of foreign object damage (FOD) ingestion during operations on the ground. This concept provides a highly embedded alternative to the Boeing baseline and is represented in Figure 3.5.
28 'D' Upper Inlet Upper BL removal
Propulsor BWB Airframe
sh Inlet w/ Internal Cavity Lower BL removal
Figure 3.5: Upper 'D' Inlet & Lower 'Flush' Inlet Schematic
Configuration 4: Aft-Fan Turbofan with Upper and Lower 'D' Inlet Here a non-traditional propulsion system is envisioned as the basis for the concept. An aft-fan turbofan would be coupled with 'D' inlets as described previously. This configuration has the same issues as configuration 3 in regards to tail skid and FOD ingestion. In addition, no engines with aft fans have been built with bypass ratios and diameters of the order required for this application. Given that the technology is not mature the concept was not pursued further. However, it represents a level of novelty so was included as illustration but was not evaluated for down-select purposes. The configuration schematic is contained in Figure 3.6.
29 BWB Airframe
Propulsor
Upper BL removal
Lower BL removal
Figure 3.6: Aft Fan Turbofan Schemetic
3.3 Configuration Assessment & Down-Select
Given the notional configurations that result from the brainstorming sessions it is necessary to implement some ordered process in which the concepts are compared and therefore eventually lead to a preferred concept for analysis. What is needed is some method to rank or score each concept against a baseline, thereby providing a metric from which to base a down-select process. Consequently, a Pugh Matrix is chosen as the tool to accomplish this goal. A Pugh Matrix is a method to compare several design ideas or configurations against a baseline using comparison criteria. The method is implemented using a tabular format as illustrated below in Table 3.1
30 Configurations
00
0 0 0
U 0 0 0
Comparison Criteria ) 0 0 0 z0 z 0 z0
* Criteria A + - S 0 Criteria B - + S 0 Criteria C - -_- * Criteria D + + S * Criteria E + + S e Criteria F S ++ - 0 Criteria G ++ - -- 9 Criteria H - S +
(S): Same (+): Better than base (-): Worse than base
Table 3.1: Pugh Matrix Example
In order to score or rank each of the concepts a method of +'s and -'s is implemented. For each criterion the candidate configurations are compared to the benchmark pod-pylon configuration. If the concept is better than the baseline it receives a plus, worse a minus or if it is the same an 'S' is used. Multiple +'s and -'s are used to provide higher levels of fidelity for comparison. The total number of marks is summed vertically for each column producing an indicator of the preferred concept; the greatest number of plus signs indicates the best performing configuration. The foundation of the technique is the information used for scoring the matrix since it will ultimately determine the preferred concept. Wherever possible analytical methods are used including equations, charts and historical data to base the evaluations. However, oftentimes Delphic processes must be used which essentially implies relying on the good judgment of experienced individuals to determine the matrix scoring. The power of the technique is that it allows a concise
31 visual representation of the positive and negative attributes of a collection of concepts. In this way it facilitates the down-select process quite efficiently. Here no weightings were assigned to the criteria but could be added to emphasize the importance of one criterion over another.
Central to the Pugh Matrix technique is the comparison criteria used to evaluate the various concepts. This criteria determines the basis for configuration validity so it is essential that the list is complete and relevant. For the BLI configuration down-select process, the comparison criteria are decomposed into two groups, those concentrating on Performance and those concentrating on Safety & Cost. The two groups of criteria are listed in the following table:
Pug:h Matrix Comparison Criteria Performance Criteria Inlet Distortion Torpedo Effect / Profile Drag Reduction Cycle Efficiency Cruise SFC Drag - Lift Induced Drag Wetted Area Drag Drag - Trim Drag TOGW Drag - Interference Drag Safety & Cost (Operation & Acquisition) Criteria
Operability Maintenance - Labor Engine Burst Considerations Maintenance - Materials Foreign Object Damage Maintenance - Support Aircraft Egress - Reverser Placement High Cycle Fatigue - Vibration Manufacturability - Airframe Noise Manufacturability - Engine Table 3.2: Pugh Matrix Comparison Criteria
With the elements of the Pugh Matrix in hand the process of scoring the configurations begins. For this a combination of analytical resources and expert advice is sought. During
32 this period the matrix evaluations undergo scrutiny by the design team in an effort to ensure that the proper assessment is given to each element of the matrix. To this end an explanation is generated for each criterion to articulate the logic behind the related score. The final version of the Pugh Matrix is contained in the following pages. Table 3.3 is the Pugh Matrix corresponding to the Performance criteria and Table 3.4 is the Pugh Matrix corresponding to the Safety and Cost criteria.
33 Table 3.3: Pugh Matrix -Performance Configurations
Pylon/Pod Configuration (Base)
Boeing 'D' Inlet w/ upper BL removal
Upper 'D' inlet and lower 'Flush' inlet Comparison Criteria Upper and Lower 'D' Inlets
S Distortion -- The imbedded concept will have more distortion due to the mixing of boundary layer and free stream flow.
* Cycle Efficiency (pressure The imbedded concept's pressure loss from the free stream flow in the boundary - -- -- layer will cause the thermal efficiency of the core to be lower than that of the recovery) baseline engine.
The imbedded concepts will have propulsion-induced circulation or load resulting in " Drag - Lift Induced Drag S S S wing span load differences from ellipticcan be addressed with wing twist or camber design changes hence keeping the lift-induced drag the same as the baseline.
* Drag -Trim Drag + + + The imbedded concept reduces the moment arm produced from the pylon/pod configuration, which reduces the amount of elevon needed to trim the aircraft.
" Drag - Interference Drag + + ++ + The imbedded concepts have few intersecting surfaces such as the pylon-wing juncture and pylon-nacelle juncture, giving it lower interference drag.
" Effect / Drag Reduction + + ++ +++ The imbedded concept will ingest the upper/lower surface boundary layers, Torpedo decreasing the aircraft's overall drag theoretically.
Cruise SFC The imbedded concept will have reduced SFC due to thermal efficiency delta as well as the distortion influence onturbo machinery performance.
" Wetted Area Drag - Cd + ++ + The imbedded concept will have less wetted area drag because it is more imbedded. This is left "to be determined" because it is integrative and dependent on many of these factors listed here. But the logic of imbedded and decreasing the number of * TOGW ? ? ? parts should decrease the TOGW. But there is another side to this logic, by imbedding the engines, it maybe be necessary to increase the engine size, thus increasing the TOGW.
34 Table 3.4:Pugh Matrix -Safety & Cost (Acquisition& Operations) Configurations
Pylon/Pod Configuration (Base)
Boeing 'D' Inlet w/ upper BL removal
Upper 'D' inlet and lower 'Flush' inlet
Comparison Criteria Upper and Lower 'D' Inlets
* Operability ------The imbedded concept will havereduced stall margin owing to inlet distortion.
critical structural and mechanical " Burst Considerations - - - The imbedded concepts place the engines closer to Engine components thus needing more structure or material to protect it making it heavier. The imbedded concepts ingestingthe lower surface boundary layer increase the risk " Foreign Object Damage + -- -- of FOD during takeoff and landing (runway debris). But by placing the engines lower, the chance of ingesting a bird is lower. * Aircraft Egress - Reverser The imbedded engine placement may create a hotter region aft due to the close Placement proximity.
" Manufacturability (airframe) - - - The imbedded concepts are more integrated leading to fewer parts to manufacture.
* Manufacturability (engine) S S S Traditional turbofans withnominal levels of technology will be considered.
are highly integrated which may require more time for * Maintenance - Labor ± - - The imbedded concepts access and repairs.
the same materials for maintenance but there " - Materials + + + The imbedded concepts will require Maintenance will be fewer parts.
may require additional support to address life cycle issues " Maintenance - Support The imbedded concepts (see next).
S High Cycle Fatigue (fan) -- The imbeddedconcepts have increasedinlet distortionwhich will presumably g y gue (f) degrade the life of the fan blades more than the fan of a pylon/pod configuration.
soundproofing (insulation) as well as " Noise + ++ ++ The imbedded concepts allow for additional more positive reflection of fan / turbo machinery noise.
35 With the scored Pugh Matrix in hand the process of generating the preferred concept commences. Essentially, the plusses and minus are summed for each column and the configurations are compared with the most positive score representing the final preferred concept. The result of the scoring process is Configuration 4, upper 'D' and lower 'flush' inlet, being down-selected as the preferred concept.
3.4 Boeing Feedback
Upon completing the process of configuration generation and down-select a packet is prepared for Boeing which outlined the procedure and summarized the results. The objective is to get expert feedback on the ideas for the novel propulsion integration concept. A letter is drafted with the particulars of the philosophy and an explanation of the candidate concepts (Appendix 3). The Boeing package included:
1. Final down-select Pugh matrix with scores 2. Highlighted chosen preferred concept (Upper 'D' & lower 'flush' inlets) 3. Letter with detailed explanations on the process
The response was very positive with interest expressed in the lower flush inlet concept. Boeing provided some insight on the Pugh Matrix including some minor modifications.
With the concept in hand the analysis on the engine and aircraft proceeds. The remainder of this thesis concentrates on the engine analysis and design ramifications.
36 4. Parametric Cycle Analysis
For the engine subsystem the predominant impact of BLI on engine performance is the associated reduction in pressure recovery that the propulsion system is subjected to [8]. It is this parameter which will have the greatest impact on the engine performance and as such this influence must be quantified. To this end a parametric cycle analysis is conducted that serves to investigate the cycle performance detriment owing to the loss in pressure recovery and the related size implications for the propulsion system. However, before discussing the results, it is instructive to review some fundamentals of propulsion system analysis. For this, simple actuator disk theory is used to describe the physics involved.
4.1 Fundamental Propulsion Theory
The thrust equation resulting from the simplified momentum equation in control volume form can be written as:
PhdSJT =-T=(V -V, ) (Eq. 4-1) m
Here the thrust, T, can be expressed as either the integrated sum of the internal pressure forces or equal to the change in the momentum flux of the fluid across the control volume. Both perspectives can yield interesting insight and both will be treated in turn.
Consider the idealized propulsion system in Figure 4.1:
37 Propulsor PT2/PT1 = RPR
Thrust ~ V2 - V.
Figure 4.1: Idealized Propulsor
The above figure represents an actuator or a propulsive disk. This can be envisioned as a propeller or fan, but the analysis can be applied equally well to a turbojet or turbofan. A fan acts to increase the momentum of the flow via the transfer of mechanical work which comes in the form of a stagnation pressure ratio (Pr2/PTo) of the gas across the disk, referred to hereafter as the rotor pressure ratio (RPR). Higher stagnation pressure results in an increase in the velocity of the gas as it expands to ambient conditions downstream. Assuming incompressible flow the Bernoulli equation can be used to calculate the downstream velocity, V2.
P pV2 (Eq 4-2) P = PS + 2 (Eq. 4-3)
V2 = 2 [ RPR * PT -Pa]
The higher velocity represents an increase in the axial momentum and consequently a thrust in the forward direction.
38 T = pVA,(V 2 -Vo)= rhAV ( Eq. 4-4)
Here it is clear that specific thrust (thrust/unit mass flow) is proportional to the velocity increase across the control volume which implies total thrust scales with airflow.
From the perspective of pressure forces, the thrust is equal to the static pressure differential across the disk multiplied by the area of the disk. Since the velocity of the gas is constant across the disk (to satisfy continuity) the Bernoulli equation says that the difference in static pressure is equivalent to the difference in stagnation pressure or:
T = A, (PT2 - P ) (Eq.4-5) T = ApPT(RPR -1) (Eq. 4-6)
Here, thrust is proportional to the disk area, Ap, which as shown above is related to the mass flow rate through the propulsion system. The link between the two perspectives is the airflow - area link and will become useful when analyzing the impact of upstream stagnation pressure loss which follows below.
Consider the same situation but with the addition of a stagnation pressure loss mechanism upstream of the propulsor, as can be seen in Figure 4.2:
39 Control Volume
V0 V, PTO P V2 P~ P __ PT2
Loss Mechanism - T Thrust ~ V2 -VI Figure 4.2: Actuator Disk with Stagnation Pressure Loss Let ni represent the pressure ratio across the loss mechanism which by definition will be less than one. For the new situation equation 4-3 is rewritten to account for the loss in stagnation pressure and therefore determine the impact on thrust. The modified equations are as follows: P'p <(Eq. 4-7) PT = r i * PrO V2 = 2 (RPR * ir,* Pr - P, (Eq. 4-8) P Clearly, from the above relation, as the pressure loss upstream is introduced the downstream velocity is reduced from the ideal (no stagnation pressure loss) case for the same rotor pressure ratio. The reduction in the downstream velocity represents a reduced momentum flux and therefore reduced thrust. Figure 4.3 illustrates the impact on downstream velocity, and therefore specific thrust, as a function of pressure recovery (ni). 40 V2/V2: 0.9 0.8 0.7 0.6 0.5 0.4 Pressure Recovery (PT1IPu) Figure 4.3: Influence of Pressure Recovery on Downstream Velocity Given the reduction of specific thrust owing to the stagnation pressure loss, the propulsor must grow in size in order to maintain the same overall thrust level. This growth implies an increase in the airflow ingested by the propulsor if the RPR is kept constant. In order for the airflow to increase the disk area of the propulsor must increase as well. Consequently, the overall impact due to an upstream pressure loss would be a larger propulsor for the same thrust level. Revisiting the perspective of pressure forces, the same conclusions can be drawn. Specifically, the upstream pressure loss would be manifested in a reduced PT2 (for same RPR). The result is a lower AP across the disk and hence a reduced thrust. Here, to increase the thrust to the original level the disk area would need to be increased. Therein lays the connectivity alluded to earlier between the two perspectives. That is whether one considers the momentum change across the control volume or the pressure forces acting on the actuator disk the same conclusion is drawn. In the former the increased airflow requirement would drive the disk area through the pViAp dependence. In the latter 41 instance the area increase is a direct consequence of AP*Ap. This connection, while not essential, provides a deeper understanding of the principles involved. 4.2 Parametric Cycle Results for Turbofan Engines With the above foundation one can now investigate the impact of pressure recovery on mixed-flow turbofan engines. To this end, a parametric cycle analysis is performed on a mixed-flow turbofan cycle similar to that proposed for the BWB propulsion system. The cycle analysis serves to illustrate the impact of pressure recovery on the fundamental performance and design characteristics of the engine including specific thrust, fuel consumption and propulsor sizing (for a specified thrust level). The process was carried out using the Pratt & Whitney tool SOAPP (State-of-the-Art Performance Program). SOAPP is an engine design and analysis tool very similar in scope to NASA's NPSS (Numerical Propulsion System Simulation). The model engine for the analysis is a mixed-flow, twin spool turbofan. A schematic of such an engine follows in Figure 4.4. All analysis was performed on-design, which implies looking at a rubber engine, at the reference flight condition of 35000 feet and 0.85 Mach. For this study all cycles are designed to the same thrust level therefore illustrating the influence on propulsor sizing as a function of inlet recovery. The design point cycle summary is contained in Table 4.1: 42 Fan Low Pressure Compressor High Pressure Turbine Low Pressure Turbine High Pressure Compressor Combustor Figure 4.4: Fan-Lo-High Turbofan Schematic Mixed Flow Turbofan Design Point Cycle Summary Fan Pressure Ratio 1.6 LPC Pressure ratio 2.5 HPC Pressure Ratio 20 Maximum Turbine Inlet Temperature (F) 2540 PT Ratio 1.1 Net Thrust 12000 lbs. Table 4.1: Cycle Design Point The study is conducted by varying the inlet recovery (ni) from ideal recovery of 1.0 to some arbitrary reduced recovery of 0.8. This study yields performance curves, trends, and influence coefficients useful for later portions of the project. Equally as important it gives insight into the physics behind the process and a feel for the magnitude of the changes. 43 4.2.1 Engine Specific Thrust and Airflow Demand As in the idealized example discussed above, the impact of reduced pressure recovery is a reduction in the specific thrust and an increase in airflow for a given cycle and thrust level. The purpose of the cycle study was to determine the magnitude of the performance debit. Figure 4.5 illustrates the results of the cycle study: 1 ------...... -----.--- 2.5 2.3 U 0 0 0.75- Design Point Cycle Summary FRR = 1.6 --2.1 LC PR = 2.-- Edt Jet Velocity Ui HPC PR =20g 0. 1...... -...... - ...... HPUPF-2 0 ...... 0 T4=254------R R 4) PT Ratio = 1.1 N 1.9 Z0 0 Z 0.25 1.7 0 4- -4 1.5 0.75 0.8 0.85 0.9 0.95 1.05 Inlet Recovery Figure 4.5: Impact of Pressure Recovery on Nozzle Exit Velocity Above it is clear that a significant reduction in the exit velocity of the propulsion system is experienced as the pressure recovery decreases. For this cycle the impact is about 6% reduction in nozzle exit velocity for an 8% reduction in recovery. The associated impact on the cycle specific thrust and airflow can be seen in Figure 4.6. 44 16 14 12 1.65-E Design Point Cycle Summary 01 FPR =1.41 0 LPC PR r2.6 CE HPC PRIF 20 T4F= 2640 -- Total Airlow PT Ratio= 1.1 --- Specific Thrust * 1.4...... -...... K tlijf 20 Ib ...... 0 C0 Z C -4 -2 0.9 0 0.75 0.8 0.85 0.9 0.95 1 1.05 1.1 Inlet Recovery Figure 4.6: Airflow & Specific Thrust Delta's Here, the ramification of the lower exit velocity is shown. As a consequence of the reduced momentum flux, the cycle output, specific thrust, is greatly reduced. Consequently, the airflow requirement to maintain the same thrust level must increase to overcome this detriment. The net result of the increased airflow is an increase in the size of the propulsion system, specifically the fan diameter. This will be discussed in the following section. 4.2.2 Fan Diameter Sizing The increased airflow need resulting from the reduced specific thrust serves to influence the size of the fan. This is the result of the aerodynamic constraints imposed on the fan sizing. Compressors and fans are designed to handle a given flow/unit area which corresponds to a given Mach number at the inlet to the component. For efficiency reasons, the Mach number at the face of the fan is typically limited to less than 0.6. This constraint forces the diameter increase as summarized in Figure 4.7. 45 Fan diameter sizing is constrained by an inlet limiting Mach number of- 0.6 (aerodynamic limitation on the turbomachinery) which corresponds to a fan specific flow capacity of 41.5 lbm/ft2 . Consequently, fan size is dependent on total airflow i.e.: Given Airflow w/ a set Mach # corresponds to a given hole size (diameter) -- T Airflow = Velocity * Area ----- a Velocity - Mach # ----ILArea - Diam eter If velocity is constrained (limiting Mach #) increased airflow requires increasedflow area (fan diameter). Figure 4.7: Fan Diameter Sizing With this constraint Figure 4.8 contains the fan diameter sizing results for the parametric study: 46 60.00% Design Point Cycle Summary 50.00%-FPR -1.6 LPC PRa 2.5 HPC PR= 20 T4F =2540 PT Ratio- 1.1 40.00% - ...... ------...... - -- 0.....0...... b.. E 0 30.00% IL 20.00% ------...... -----...... - .-..-..---- .--... 10.00%- 0.00% 0.75 0.8 0.85 0.9 0.95 1 1.05 Inlet Recovery Figure 4.8: Fan Diameter Size Trends The increase in the required fan size for the propulsion system has significant design ramifications for the propulsion system. Foremost of these is the increased weight of the engine (reduced thrust-to-weight ratio). Other considerations include aircraft installation challenges and manufacturing difficulties of the larger geometries. 4.2.3 Overall Efficiency and Specific Fuel Consumption The overall efficiency of a gas turbine propulsion system is the product of the thermal (cycle) efficiency and the propulsive efficiency. Thermal efficiency is a measure of how effectively the thermodynamic cycle converts the thermal energy in the fuel to net cycle output. Propulsive efficiency relates how well the net cycle output is converted to produce thrust power, the useful output of the propulsion system. BLI, through the reduction in inlet recovery, affects both contributors. These influences are now expanded on. The operating cycle of a gas turbine propulsion system is the Brayton cycle. The thermal efficiency of a non-ideal Brayton cycle is primarily a function of the compressor and 47 turbine efficiencies, inlet recovery, maximum cycle temperature and overall cycle pressure ratio (Pmax/Pmin) [4]. Assuming constant component efficiency levels, compressor pressure ratio, and cycle temperature, thermal efficiency is dependent on the inlet recovery. Using simple cycle analysis one can quantify analytically the influence of BLI on the thermal efficiency. The thermal efficiency of a non-ideal Brayton cycle can be expressed as: ad adT 77c 77, - T~s 7,=(1- T) TO (Eq. 4-9) 1 + r - 1( -I The influence of BLI on cycle performance is communicated via a reduction in the adiabatic compressor efficiency, io, with the following relation: acBLI a adBLI (Eq. 4-10) -1 (Eq. 4-11) r1BLI sz~ 1if)f~1 7s Substituting equation 4-10 into equation 4-9 the effect on thermal efficiency is captured. As evidence, refer to Figure 4.9. Here the thermal efficiency of a non-ideal Brayton cycle is shown as a function of inlet recovery for nominal levels of cycle parameters. The strong negative impact of pressure recovery on thermal efficiency is clear. Therefore, BLI through the reduction in effective inlet recovery reduces the thermal efficiency of the propulsion system. 48 0.6 ------.--.- - -- .- .- .- -. 4) ~ 0.4- 0 0.5 0.6 0.7 0.8 0.9 1 Inlet Recovery Figure 4.9: Non-Ideal Brayton Cycle Thermal Efficiency Propulsive efficiency is related to the level of wasted kinetic energy trailed by a propulsion system in its wake. Consequently, the magnitude of an engine's jet velocity determines the propulsion efficiency. Ideally, the propulsor would leave zero wake behind and hence waste no energy. However, this situation would result in zero thrust so therefore some level of exhaust velocity is needed. For a single stream propulsion device, such as a mixed flow turbofan, the propulsive efficiency (71p) is quantified as follows: 2 (Eq. 4-12) 7 P=V1+ Je V, Here, for any flight velocity (V 0) propulsive efficiency is increased by reducing the jet velocity. As is shown in Figure 4.5, BLI serves to decrease the jet velocity via a reduced 49 nozzle pressure ratio. Therefore the propulsive efficiency increases as the inlet recovery is reduced. The relative impact of BLI on the thermal and propulsive efficiencies is what determines the impact on the overall efficiency for the propulsion system and hence the specific fuel consumption. To illustrate how BLI influences both efficiencies consider Figure 4.10. 20.00% -Propulsive Efficiency 10,00%- --- ThermalEfficiency 0.00%- 0. '5 0.8 0.85 09 0.95 -10.00%- 0 -20.00%- -.... . - -. . ..- -. -- -30.00%------ U.r -40.00%- -50.00% Inlet Recovery Figure 4.10: Relative Impact of Inlet Recovery on Thermal and Propulsive Efficiency Here the relative gain in propulsive efficiency is about % of the loss in thermal efficiency. More clearly, the propulsive efficiency gets better less than the thermal efficiency degrades. As a result, the overall impact is a reduction in the overall efficiency of the propulsion system. With SFC inversely proportional to overall efficiency, an increase in specific fuel consumption results. Figure 4.11 shows the TSFC and overall efficiency results from the parametric study. 50 00.00% 40.00%-- - SFC 30.00% - -- - Overall Efficiency 20.00% ------. 10.00%- 0.00%~ 0. 5 0.8 0.85 0.9 0.95 - 1W -10.00% - - .. -- - ...... -...... -...... -20.00% - --- -...... -30.00% ...... - . ...- ...... -40.00% Inlet Recovery Figure 4.11: Relative Impact of Inlet Recovery on SFC and Overall Efficiency 4.2.4 Gas Generator Core Size Impact Physically, the necessity of larger airflow for a given thrust level requires that more horsepower be created in the gas generator (core) to drive the same fan pressure ratio with greater airflow. This additional power comes from increased core airflow and hence increased fuel flow (constant combustor exit temperature). Therefore, the result of the reduced pressure recovery is a larger core to power the propulsion system. As evidence, Figure 4.12 shows the increased core size (core airflow) as a function of inlet recovery for constant thrust level. 51 70.00% 60.00% 50.00% ...... - Desig P it Cy Su y - FPR= 1.6 D LPC PR = 2.6 HPC PR = 20 40.00% ------.. ------4 E 2 M 0 PT Ratio = 1.1 o Net Thrust = 12000 lbs. o 30.00%- 20.00% - - --.--. -..------..-. . .. 10.00%- 0.00% 0.75 0.8 0.85 0.9 0.95 1 1.05 Inlet Recovery Figure 4.12: Core Size Impact of BLI 4.3 Cycle Analysis Summary Owing to BLI and the associated reduction in pressure recovery the engines have the following performance attributes with respect to a traditional pylon/pod arrangement: - Reduced Specific Thrust - Reduced Thrust to Weight Ratio * Increased Fan Diameter - Increased Fuel Consumption - Increased Core Size With these attributes it is not obvious why ingesting boundary layer air would be beneficial to the aircraft system as a whole. However, the benefit of the reduced profile drag of the aircraft has not yet been analyzed. The next section will incorporate BLI drag reducing effects in addition to these engine performance trends to investigate the possible fuel bum benefits for a BLI configuration. 52 5. Propulsive Cycle Design With the underlying performance impact of BLI in hand from the cycle study, the next portion of the project investigates the influence of BLI on the optimum propulsive cycle selection. More clearly, the question of interest is "how does the fuel burn reduction for a BLI propulsion system trend with fan pressure ratio (diameter) as compared to a baseline traditional pod/pylon installation?" The notion is that starting from some baseline traditional installation the embedded engines would provide a fuel burn reduction owing to the inherent benefits of BLI. However, when one factors in the weight reduction due to embedding and the additional profile drag reduction resulting from increased engine airflow, there may exist room for the engine's propulsive cycle to change towards lower fan pressure ratios (larger fan diameters) and therefore provide further reductions in realizable fuel bum. This then presents a close coupling between engine cycle performance and system weight with regards to fuel burn. The purpose of this section is to describe a methodology that was used to investigate that coupling and then present results consistent with that philosophy. Various factors comprise aircraft fuel burn as illustrated in the tree diagram of Figure 5.1: Fuel Burn: lbs./hr Specific Fuel Consumption Aircraft Thrust TSFC Requirement I I Engine Performance Drag Propulsive Cycle Inlet Recovery System Weight Degree of BLI Figure 5.1: Factors Comprising Fuel Burn 53 As evident in Figure 5.1, aircraft fuel burn is a function of both the aircraft and engine performance. Therefore, in order to estimate the fuel burn reduction for the notional BLI configuration both of these influences had to be included. This trade study focuses on selecting the optimum propulsive cycle with allowances for the influence of total system weight, BLI profile drag reduction and the associated propulsion system pressure recovery. Overall the design study incorporated the following: - BLI profile drag reducing effects = Embedded configuration weight benefit - BLI impact on engine performance - Fan & core size influence on propulsion system weight = System weight effects on fuel burn (through trade factors) To facilitate handling of the problem the methodology in Figure 5.2 was implemented: - Trade study carried out using an excel spreadsheet - Engine performance data generated using Pratt & Whitney SOAPP program - Boundary layer properties calculated using a 1-D flat plate analysis Engine Performance w/ -sfc Excel Spreadsheet Optimum BLI Cycle -Diameter Traet Methodoly -weight Boundary Layer Model _Inet recovery Aircraft Trade Factors - I~Weight vs. Fuel Burn I Figure 5.2: Fan Pressure Ratio Trade Study Methodology 54 Central to the analysis was the evaluation of a series of engines all designed around a common core but utilizing different propulsive cycles (fan pressure ratios). The core thermodynamic cycle was the same as that used for the parametric cycle analysis. Output from the study would be a trend of fuel burn with fan diameter resulting in the optimum propulsive cycle within the given constraints. Each of the above elements from Figure 5.2 is now expanded upon. 5.1 Boundary Layer Model Given that the foundation of the study involves the ingestion of the boundary layer flow of the BWB aircraft, a model of the flow conditions is required. For this a crude approximation of a 1-D flat plate turbulent boundary layer is used [8]. Since the interest is design trends and fuel burn estimates the crudeness of the boundary layer model seems reasonable. In addition, the upper & lower surface boundary layers of the BWB are assumed to be identical. Again, this is a simple approximation but one sufficient to illustrate the trends. The boundary layer model provides a velocity profile and a thickness at an averaged span location derived from the baseline BWB configuration geometry and using a reference flight condition of Mach 0.85 at 35000 feet. Given the profile and thickness the boundary layer mass flow (per unit depth) and average velocity is calculated. Following from this, the momentum deficit in the boundary layer representing the profile drag is calculated. The following table summarizes the key boundary layer characteristics: BWB Boundary Laver Model Characteristics BL thickness (m) 0.329 Average velocity, Vavg/Voo 0.845 BL mass flow per unit depth: (kg/s) 0.2537 BL drag per unit depth: (N) 9.8 Table 5.1: Boundary Layer Properties 55 5.2 Engine Performance The fan pressure ratio (FPR) trade study requires the generation of performance data for a range of propulsive cycles. Each FPR perturbation is defined with a common core cycle. A baseline engine cycle is defined that represents the traditional pylon/pod configuration which also shares the common core cycle. Data generation is conducted via the Pratt & Whitey thermodynamic performance tool SOAPP. The FPR sweep incorporates the same core cycle parameters as the parametric cycle analysis (see section 4). In this instance the FPR is varied from 1.5 to 2.2. All engines are sized to the same thrust level of 14786 lbs. as in the Rodriguez study [1]. All cycles were defined with an inlet recovery level of 0.95 which was an average value consistent with removal of the upper & lower surface boundary layers. The adjustment for the inlet recovery being different from this level is corrected with influence coefficients later. With these ground rules the following performance data is generated: Airflow: Core Size: Diameter: SFC: FPR | BPR lb/s Ib/s in 1/hr 1.5 9.2 1405.7 12.56 131.3 0.571 1.6 7.6 1176.9 12.50 120.2 0.570 1.7 6.4 1020.1 12.53 111.9 0.573 1.8 5.5 907.2 12.64 105.5 0.580 1.9 4.8 821.3 12.77 100.4 0.588 2.0 4.3 753.2 12.92 96.1 0.596 2.1 3.9 697.8 13.08 92.5 0.605 2.2 3.5 652.3 13.26 89.5 0.614 Table 5.2: FPR Sweep Performance Data The above metrics represent the most important parameters for selection of the optimum propulsive cycle. These parameters comprise the input to a spreadsheet analysis tool to be described later. Here a point is addressed which will be important downstream in the analysis. Examining the data in the above table one recognizes that the SFC decreases as the FPR decreases. This is the result of increasing propulsive efficiency. For all else equal, the lower the FPR (higher bypass ratio) the less fuel is consumed. Consequently, 56 when one is selecting the propulsive cycle for a given mission one would typically chose the lowest FPR that is consistent with favorable aircraft installation in terms of size, weight, and drag [10]. The baseline comparison engine was defined with a FPR of 1.7 which is consistent with a generic turbofan cycle for a traditional installation. This cycle was defined with near ideal inlet recovery as would be expected in a pylon/pod installation. The performance summary for the baseline cycle can be found in Table 5.3. Airflow: Core Size: Diameter: SFC: FPR BPR lb/s Ib/s in 1/hr 1.7 6.2 934.0 11.86 104.3 0.539 Table 5.3: Baseline Cycle Performance Data 5.3 Engine Inlet Recovery & BLI Drag Reduction Calculation The link between engine performance and profile drag reduction is the effective inlet recovery of the propulsion system. For this analysis, the effective inlet is taken as the portion of the aircraft in front of the engine (see section 2.5) over which the boundary layer develops. With the boundary layer model, the actual pressure recovery is calculated given the airflow demands of the cycle. An assumption here is that the additional airflow demand by the engine is supplied from the freestream, which when mixed with the available airflow in the boundary layer results in some loss of stagnation pressure for the propulsion system. The calculated recovery is then compared to the recovery assumed during the engine definition and the difference is corrected with trade factors. Figure 5.3 illustrates the connectivity. 57 Engine Performance Boundary Layer Model -Diameter -Available Airflow/unit depth -Airflow Demand -Drag /unit depth -SFC *BLI Profile Drag Reduction *Actual Engine Inlet Recovery Figure 5.3: Engine Inlet Recovery & BLI Drag Reduction Calculation For this analysis, the profile drag reduction due to boundary layer ingestion is presumed equal to the drag of the effective inlet or the strip of fuselage in front of the engine. The width of the strip is set equal to the diameter of the engine thereby facilitating a straightforward calculation given the boundary layer model. In effect, the drag/unit depth value of the boundary layer model is multiplied by the fan diameter (depth) to yield the total drag reduction for the configuration. In the case where both the upper and lower surface boundary layers are ingested this value is multiplied by two. In essence, this calculation models the drag of a two-sided flat plate under the given flight conditions. This drag number is then subtracted from the pod/pylon thrust requirement creating a BLI thrust requirement. The new fuel bum is then the engine SFC multiplied by this "new" thrust requirement. Total airflow available in the boundary layer for engine ingestion is estimated as the engine diameter multiplied by the airflow/unit depth from the boundary layer model. This assumes no feedback of the engine flowfield on the aircraft aerodynamics, which is not 58 accurate. However, for this level of analysis the assumption seems reasonable. Again, as in the drag reduction calculation the total airflow is comprised of both the upper and lower surfaces. With the engine airflow being comprised of both the boundary layer and the freestream flow there must be mixing of the two flows as the air enters the engine. It is this mixing which determines the level of effective pressure recovery for the propulsion system and is also responsible for the increased levels of distortion which will be discussed later. The figure below serves to illustrate the situation: Propulsion System NMNG Fan Face Inlet Boundary Layer Freestream Figure 5.4: Sources of Engine Airflow Given the cycle airflow demand from the engine performance analysis and the available airflow from the boundary layer model the effective engine inlet recovery is determined. This entails a mixing calculation of the boundary layer flow and any additional freestream flow that is necessary to satisfy the engine demand. Using a mass average technique the mixed flow velocity is computed and therefore the mixed pressure recovery follows. The resulting inlet recovery is different than the 0.95 assumed during the cycle design process. This requires that the engine performance (SFC & diameter) be corrected through trade factors to account for the difference in recovery. This feedback is illustrated 59 in Figure 5.3 as the arrow pointing back toward the engine block. The step of correcting for the actual recovery is very important given the strong influence of recovery on SFC and the importance SFC has in determining the resulting fuel burn for the configuration. 5.4 BLI Weight Reduction & Trade Factors A key contributor to the performance of a highly integrated BLI configuration is the associated potential weight savings of the concept. The largest contributor to propulsion system weight reduction from the pod/pylon configuration is the removal of the weight associated with the pylon structure that supports the engine. This is a significant piece of structure which can represent as much as 30-40 % of the weight of the engine itself [16]. Embedding can not remove all the necessary structure to hold the engine to the aircraft though it would provide for a significant reduction. For the purpose here the weight savings due to embedding is taken as 25% of the weight of the baseline pod/pylon 104.3 inch turbofan. For a bare engine weight of 12000 lbs. this represents a 3000 lb. reduction in system weight. In order to model the impact of weight changes on the performance of the system a series of trade factors are implemented which transform weight directly to fuel burn. The following trade factor obtained was used to this end: 1000 lbs. Weight = 0.82% Fuel Burn (Eq. 5-1) For the FPR trade study, a series of engines are investigated all with different geometries (fan diameter & core size) and therefore system weights. The impact is calculated with the above trade factor and a series of additional trade factors to transform fan and core size deltas into weight increments. Here the trade factors used are: 150 lbs. / inch of Fan Diameter (Eq. 5-2) 20 lbs. / % Core Size (Eq. 5-3) 60 With these two ratios the differences in system weight from the traditional turbofan in the pod/pylon configuration to the embedded engines are determined. These differences, along with the system weight - fuel burn trade factor and the lump weight savings due to removal of the pylon structure support the calculation of fuel burn reduction explicitly due to weight change. 5.5 BLI Influence on Component Performance Owing to the distortion present due to the mixing of the boundary layer and freestream flow entering the engine the performance of the turbomachinery may be adversely affected. Specifically, the polytropic efficiencies of the compression system may be reduced as a result of unsteady flow, turbulence and vorticity. Consequently, it seems logical to model this effect when calculating the fuel burn improvement for the embedded configuration. To do this a series of sensitivities is generated using SOAPP that characterizes the impact in terms of an SFC detriment. For instance, the sensitivity of SFC to a 1% reduction in polytropic efficiency is calculated for the fan and high pressure compressor. These influence coefficients are then applied to the engine performance data to simulate the effect of reduced component performance and hence illustrate the change in realized fuel burn. For illustration purposes, a plot of the sensitivity of SFC to fan stage efficiency is shown in Figure 5.5. 61 3.5% 3,0% - 2.5%- LL 2.0%- 0) 1.5% - 1.0% 0.5%- 0.0% - 0.89 0.9 0.91 0.92 0.93 0.94 0.95 0.96 Fan Stage Efficiency Figure 5.5: SFC Sensitivity to Fan Efficiency As is evident in the figure the effect is linear and strong with a one point reduction in fan efficiency causing a ~0.7% increase in SFC. Table 5.4 summarizes the influence coefficients used in the model. Turbofan Component Sensitivities -1 Point Fan Efficiency +0.69% SFC -1 Point HPC Efficiency +0.64% SFC -1 Point LPC Efficiency +0.29% SFC Table 5.4: Turbofan Component Sensitivities 5.6 FPR Trade Study Implementation Tool The process of collating all the elements of the trade study is facilitated with a spreadsheet tool. Here the engine performance data, profile drag reduction calculation, system weight trade factors and engine influence coefficients are brought together. Figure 5.6 is an example of the tool. 62 FanDiameter Study for BVE PropulsionSystem Note:Each cycle is definedwihsamethrusteqimnis& pressue rmcoverynominal es for a"base"'BUcase. Diffemncesint u df(Airfo) willbe capturd wih fuelbum Influencecoelficients wil bedveloped for ccon ConoeDela's CycleSizig Criteri Trd 50/ Oldb Takof TrI 20.C T4F 2540 WA 41.5 FNT 14786 Il-pc 00/ CPR 50.C Weigt TradeFactois #/id1Diameler 150 9#/%coresize 20 ,AWaift Swtch 1 A Veaift (Inteddng) -30 EngineFeformance Data SystemVight Influence FuelBurn Calculation Arnow ADag "NeW FPR(dimb) BPR Ibs Core Size aweter- in SFr 1Ar A KIbt (wtting) A ftWgt(engne) A \Wit (tota) (B) hmstreq Ft'Lin:IWr %Delta FB 1.5 9.19 1406 1256 131.3 0.562 -3000 1.6 7.57 1177 1250 120.2 0.561 -3000 1.7 6.41 102D 12.53 111.9 0.564 -3000 1.8 5.53 907 1264 105.5 0.571 -3000 1.9 4.85 821 1277 100.4 0.578 -3000 20 4.30 753 1292 96.1 0.586 -3000 2.1 3.85 698 13.08 92.5 0.595 -3000 22 3.48 652 13.25 89.5 0.604 -3000 Figure 5.6: FPR Trade Study Tool 63 5.7 Trade Study Results and Discussion As presented in Figure 5.2 the trade study brings together the individual elements and in turn illustrates the trend of fuel bum reduction with engine cycle. Before presenting the results it is instructive to review the basis of the study in terms of what it does and does not represent. The fan pressure ratio trade study illustrates the fuel burn trend as a function of propulsive cycle for a notional BWB configuration with boundary layer ingestion of the upper and lower centerbody. The study incorporates: = Profile drag reduction owing to BLI " Weight benefit owing to embedding = Component efficiency reductions due to inlet distortion The study does not take into account the following attributes: - Possible trim drag benefit due to embedding = Feedback influence of engine flowfield on aircraft aerodynamics - Profile drag reductions due to embedding resulting from less wetted area - Thrust requirement changes from profile drag reduction With the above in mind Figure 5.7 represents the results of the FPR study. The data is plotted in terms of engine fan diameter which is inversely proportional to the cycle fan pressure ratio. 64 6.00 4.00 2.00 E = 0.00 8 U- O -2.00 -4.00 -6.00 -8.00 Fan Diameter (in.) --- Best Case -- Best Case w/ component losses No weight benefit w/ component losses Figure 5.7: Fuel Burn Trade Study Results 65 In Figure 5.7 the reality of the fuel burn benefit is illustrated. One sees quite clearly that the fuel bum for the BLI configuration is significantly less than the traditional pylon/pod, on the order of 3-5%. Here two sets of data are presented which serve to frame the feasible solution space. The first set, which corresponds to the solid line, represents the best case fuel burn reduction where no component efficiently losses are assumed. The minimum fuel bum for this set is ~ 5.8%. The dashed set of data assumes a one point reduction in the fan and HPC polytropic efficiencies which results in maximum ~ 4.2% reduction in realizable fuel burn. Also shown on the plot is a line at 121" fan diameter which represents the maximum allowed fan diameter to keep the total system weight the same as the pylon/pod. This constraint is related to a center-of-gravity limitation that can not to be violated. Upon examination of the results the most notable attribute is the trend that indicates the optimum engine size is towards larger fans to capture the increasing benefit of BLI, until the point where the additional engine weight outpaces the drag reduction benefit and the net fuel burn begins to increase. Fortunately, the embedded configuration allows for larger fans due to the associated weight savings. Consequently, the same or more ideal propulsive cycle can be implemented with additional fuel bum reductions. For example, embedding the baseline 104" propulsion system would result in a 3% reduction in the realized fuel burn. This is indicated as point #1 in the figure. However, owing to the loss in pressure recovery the same 104" engine when embedded would not correspond to the same propulsive cycle (fan pressure ratio) as the baseline configuration for the same thrust level. The embedded 104" cycle requires a higher fan pressure ratio to meet the trust requirement and therefore has a lower propulsive efficiency and higher inherent SFC. Given the weight reduction from the embedded concept, the engine size (fan diameter) grows to allow the same propulsive cycle to be implemented. This then provides for the BLI drag reduction in concert with the same propulsive efficiency, therefore improving the fuel burn. This is indicated as point #2 in the figure. In this case an additional 1% of fuel bum could be garnered but at the expense of 7" of fan diameter with the "new" engine utilizing a 112" fan. Extending this argument further, the full weight benefit of the embedded configuration could be exploited, thereby maximizing the 66 potential propulsive efficiency through lower fan pressure ratio cycles. This case is represented with point #3 in the figure. Here a 120" turbofan with a 1.6 FPR would provide an additional 0.2% fuel bum savings. This propulsion system uses up nearly all the available weight savings from the embedding as is evident with the proximity to the weight-cg constraint. This represents the optimum propulsive cycle for the configuration utilizing BLI. Table 5.5 summarizes the attributes for the traditional and optimum embedded propulsion systems. Traditional Cycle BLI Cycle (optimum) Fan Diameter (inches) 104 120 FPR 1.7 1.6 Inlet Recovery (Pt2/PtO) 0.99 0.962 SFC (1/hr) 0.539 0.570 % Delta Fuel Burn Reduction --- -4.2% Table 5.5: Propulsion System Attributes Clearly, from the preceding analysis the propulsion system changes when the effects of embedding are fully captured. Primarily the system weight availability due to the more integrated configuration provides for the changes, but the benefits are realized even in lieu of any such provisions. The above results represent fundamental propulsive cycle and performance trends and illustrate the notion that engines designed for traditional installations would not be optimal for BLI concepts. That is to say embedded propulsion systems will tend to be larger with lower FPR cycles than their traditional modular counterparts. With the fuel bum benefit and engine cycle selection trends established the attention is now turned to the more negative attributes of an embedded propulsion system. Specifically, the compression system design challenges owing to the elevated distortion levels. This issue is treated in the next section. 67 6. Compression System Design Implications 6.1 Introduction A foremost consideration in the design of gas turbine engines is ensuring the stability of the compression system. Here, the compression system refers to the fan, low-pressure and high-pressure compressors. An example of a model compression system, similar to the type considered in this project, can be seen in Figure 6.1. Fan LPC HPC Figure 6.1: Notional Compression System Compression system instabilities, commonly referred to as surge, can have detrimental effects on the mechanical soundness of the engine and can pose serious flight safety issues for the aircraft through a loss of power and control. The problem of providing sufficient surge margin to ensure stable operation throughout the flight envelope is a difficult one, with ramifications including but not limited to, engine weight and fuel consumption. Many good sources exist that thoroughly treat the physics of compressor 68 surge [9], [10], [12], [13], [18]. The reader is urged to consult those for a more rigorous explanation. The intent here is to solely give an overview of the fundamentals. The region of stable engine operation can be visualized with the help of compressor maps and surge lines. The compressor map illustrates the operating regime of the compressor in terms of flow rate and pressure ratio with speed lines and efficiency "islands" overlaid. A surge line is also included which represents the limit of stable operation, above which instability ensues. Figure 6.2 illustrates these concepts for a generic compressor: 30 25- Surge Line 20- 0 cc S 15 ...... -.... - ...... - CL Speed Lines Efficiency Islands 5 ...... -...... -... ..-...... 0 1 40 50 60 70 80 90 100 110 % Design Corrected Flow Figure 6.2: Generic Compressor Map Representation Here the operating line represents the locus of possible operating points of the gas generator satisfying both conservation of energy and mass. Surge margin (SM), as indicated, is the difference between the operating line and the surge line and is defined as: SM = PRSurgeLine -PROperating Line @ fixed flow (Eq. 6-1) PRoperaing Line 69 The surge line is not fixed and moves throughout the operation of the engine. Several factors contribute to this, ranging from internal mechanical tolerances and transient effects to Reynolds number effects and external flow disturbances upstream of the compressor inlet [10]. The latter of these is expanded upon here as it represents the most relevant aspect for a BLI configuration. Specifically, inlet distortion owing to BLI contributes to a loss in available surge margin as the surge line moves down, towards the operating line. Distortion is by definition a region of the flow with a lower stagnation pressure. Here distortion is the result of the incomplete mixing of the freestream and boundary layer flow comprising the airflow demand of the engine. The resulting total pressure distortion is essentially an axial velocity distortion as Pt ~ pV 2/2. Consequently, the impact on the compression system is manifested through the resulting angle-of-attack changes on the compressor blading. More clearly, for a fixed rotor speed, varying the axial velocity (V) changes the angle-of-attack of the blade and hence the blade loading. In the extreme, the blade is driven to stall if the flow incidence angle becomes too large. This is illustrated in Figure 6.3: Compressor Airfoil I Blade Loading ~ C,/U Blade Speed, U = constant Axial (Absolute) Velocity C F Figure 6.3: Velocity Diagram 70 For the overall rotor assembly in a distorted flow environment, some blades will do more work than others and some may stall thus leading to instability of the compressor. The instability, compressor surge, is characterized by a complete breakdown of the pumping ability of the compressor with associated power loss of the engine. In general, this constitutes a highly unfavorable situation. For the notional BLI configuration being discussed here the intent is to obtain an estimate of the inlet distortion and the resulting magnitude of the surge margin loss. With this, possible design changes to the compression system can be explored so as to maintain the nominal surge margin and ensure safe, reliable operation. 6.2 Quantification of Inlet Distortion Effects on Stability In order to quantify the impact of pressure distortion, one must first make an estimate of the magnitude of the distortion that is present. Distortion is traditionally characterized in terms of a series of indices that reflect the degree to which the stagnation pressure departs from the average level. One such index is termed the DC (6) index and is defined as: S3600 - Worst G DC(8)= y (Eq. 6-2) - pCi 2 Here the angular sector with the lowest total pressure is chosen to determine the stability index. Empirically it can be shown that a critical sector angle exists where the impact on compressor performance is greatest [12]. With the DC level determined empirical data can be used to estimate the loss in compressor delivery pressure at the stability limit and therefore the loss in surge margin. Reid [17] has compiled data in this regard and this source is used in the proceeding analysis. As an example of the type of information available consider Figure 6.4. 71 101 100 99 - 9c DC =0.29 C 98- *95- 00 0 50 100 150 200 250 300 350 400 Angle of Spoiled Sector (dog) Figure 6.4: Spoiled Sector Angle Influence on Stability (from Longley and Greitzer [12]) Figure 6.4 illustrates the loss in surge pressure ratio as a function of the angular width of the distorted region. The data is for a given DC level and therefore characterizes the level or severity of a particular experimental condition. Studying the figure one realizes that a convenient first order estimation of the distortion impact can be obtained via characterizing the geometry of the distortion that will be expected. That along with an estimation of the DC level for the distortion allows for the calculation of surge margin loss. This is the philosophy applied to the analysis of the BLI distortion problem. Consider the installation of the proposed BLI propulsion system ingesting the upper and lower surface boundary layers. Figure 6.5 provides a sense of the geometry in question: 72 ------'D' Inlet Propulsion System BWB Centerbody Fan Face 'Flush' Inlet Figure 6.5: Inlet Configurations Schematic Here the use of the upper 'D' inlet and lower flush inlet is shown. Upon examination of this situation several key observations are made. Foremost is the asymmetry in the inlet configuration, stemming from the two distinct inlets utilized to feed the engine. This asymmetry with incomplete mixing results in different flowfields entering the inlet from the top and bottom. In addition, the upper and lower surface boundary layers have different properties (i.e. velocity profiles, available airflow, etc.). These two influences coalesce to create a net distortion pattern which is approximated by a 1800 spoiled sector. In essence, the compression system is treated as a parallel compressor pumping fluid from two separate reservoirs. More clearly, whereas each of these two separate reservoirs contains a certain level of distortion, the relative effect resembles that of a parallel compressor. Therefore, while the actual flowfield entering the fan is very complex, the fundamental differences owing to asymmetry comprise the most significant first order driver for stability. 73 With the geometry determined, the DC level for the installation is estimated. Using the boundary layer characteristics and engine airflow demand from section 5 a first order estimate is made. The boundary layer flow and freestream flow velocities are mass averaged to determine the mixed Mach number of the distorted flow into the engine. The corresponding reduced stagnation pressure is calculated and the average total pressure entering the engine is calculated with the inlet recovery from earlier. Applying the definition of the DC parameter determines the level of distortion. For this application the DC level is about 0.27. In order to use the empirical data of Figure 6.4 the DC level must be similar. As is expected, the surge delivery pressure is almost directly proportional to the distortion intensity for a fixed pattern of spoiling [17]. In this instance the two DC levels are close with the experimental results corresponding to a DC of 0.287. The two values differ by about 6%. Fortunately this is close enough to justify application without scaling for this level of analysis. Therefore the empirical data is used as is. Applying the empirical data with the assumed distortion characterization one arrives at the following conclusion for the loss of stability owing to BLI . Specifically, owing to the asymmetry and resulting "effective" parallel compressor the loss of surge pressure ratio is about 10%, a significant number. Going forward, the assumption is that the distortion is communicated throughout the compression system with no change in character. Consequently, this stability loss is transmitted to both the fan and high-pressure compressor. Therefore, the design of both of these components will need to account for this loss in surge margin. Those considerations are treated in the next section. 6.3 HPC Design Considerations With the 10% loss in surge pressure ratio identified, the first step is to calculate the loss in surge margin. The notional high-pressure compressor identified in the cycle calculations had a pressure ratio of 20:1. Applying the traditional commercial surge margin of 20% would yield a pressure ratio at surge of 24:1. However, with the 10% loss 74 in surge pressure ratio the effect is a compressor operating at a pressure ratio of only 21.6 at the stability limit. This corresponds to a surge margin of only 8%, or 60% loss in surge margin. Clearly this is an unacceptable situation that demands design changes to the HPC to restore the necessary stability margin. Figure 6.6 illustrates the stability boundary migration: 30 25 -- 8% SI 20 ------CleanSurge-Une-- - - - StabilityBoundary Migration..*- Une 5-Compressor Operating 0 40 50 60 70 80 90 100 110 % Design Corrected Flow Figure 6.6: HPC Compressor Map with Distortion For the high-pressure compressor, the most straightforward methods to increase the surge margin are to either reduce the stage loading of the entire compressor, thereby increasing the number of stages required, or drop the compressor operating line. Here the former method is sought as it maintains the cycle compression ratio and hence thermal efficiency. One way to conceptualize this is to consider adding additional SM on top of the clean surge line. Therefore, when in the distorted environment the stability limit moves down, the migration starts from a higher level thus leaving the required margin (20%) after the travel. The compressor is designed to produce the new surge pressure 75 ratio with the "old" compressor's maximum stage loadings, thus the need for more stages. In this way the impact of distortion is accounted for in the compressor design with an assumed constant level of technology to facilitate a viable comparison. It is assumed that the original compressor (no distortion) produces a pressure ratio of 20:1 in ten stages. The surge pressure ratio is therefore 24:1. The surge stage loading (average stage pressure ratio) is then 24(") = 1.374. In order to produce a pressure ratio of 26.7 10.5 stages of compression are required. Consequently the need for one additional compressor stage is clear. The design philosophy is summarized in Figure 6.7 below. 30 Margin 25 20 0 it 15 (0 I 0. 0 10 Compressor Operating Line 5- 0-4- 40 50 60 70 80 90 100 110 % Design Corrected Flow Figure 6.7: HPC Stability Re-Design Here it must be mentioned that this is a very simplified analysis which does not take into account the effects that stage matching has on the compressor's stability characteristics. Stage matching is a non-trivial issue which as strong implications on the compressor performance and stability. 76 Table 6.1 summarizes some of the design characteristics for a traditional (pod/pylon) and distorted (BLI) high-pressure compressor: Nominal HPC Distorted HPC Pressure Ratio 20:1 20:1 Surge Margin, % 20 20 Stages 10 11 % Delta Weight + 5 Table 6.1: HPC Design Summary Reviewing these results it becomes clear that the high-pressure compressor design will need to accommodate the surge margin loss. The additional stage represents a weight and length increase over the traditional (pod/pylon) design. 6.4 Fan Design Considerations The fan design often represents the most critical design challenge for high bypass ratio turbofans. While the HPC carries most of the cycle compression ratio, the fan delivers the majority (-70%) of the engine thrust and therefore its importance for the engine performance cannot be overstated. As in the high compressor design, the fan will lose a significant amount of surge margin owing to the presence of the large levels of distortion. However, given that the fan is comprised of a single stage, the ability to add additional stages to replace the stability margin is not an option. Instead, the fan will have to regain surge margin primarily with an increase in tip speed and less with aerodynamic changes such as solidity (i.e. number of and spacing of blades). Here the analysis is applied to an isolated fan stage with the system aspects of stability not considered. Also, the downstream impacts of the fan design on the LPC and HPC are not accounted for. The notional cycle for the proposed embedded propulsion system has a 1.6 fan pressure ratio. Assuming again 20% surge margin, the stalling pressure ratio for the fan (at constant flow) would be 1.92. BLI distortion will contribute a 10% loss in surge pressure 77 ratio, which implies the effective "clean" surge line must be increased to a pressure ratio of 2.11 in order to provide sufficient viable margin. The increase in stalling pressure ratio must be accomplished with a speed increase of the fan. To see why this is so, consider Equation 6-3: Ah T = 2h (Eq. 6-3) U2 Here U is the fan tip speed and y represents the stage loading of the rotor which represents an aerodynamic constraint that for a fixed technology level can be assumed constant. As a consequence, fan speed increase becomes essential for increased pressure ratio (Ah) [4]. Using a numerical model of the fan within the SOAPP tool, one can obtain an estimate for the speed increase required to increase the surge pressure ratio. Specifically, a 10% increase in FPR at constant flow would require a 4% increase in fan tip speed. It seems safe to assume that a speed increase of similar magnitude would be required for the BLI engine to maintain stall margin. The speed increase will not come without a price. Most importantly, the low spool has to increase in mass (weight) in order to absorb the increased centrifugal loads of the higher speed fan. Also, the fan efficiency may be less owing to increased shock losses from the higher tip speeds and non- uniformities along the span of the blades [9]. Expanding upon the last point, when one analyzes the BLI distortion problem it becomes quite apparent that the problem is in a sense, steady state distortion. This is quite different from the more usual case where the distortion is the result of transient phenomena, such as a maneuver. Because of this perhaps the possibility would exist to change the stagger on the blading so as to more optimally receive the inlet flow. In this way some of the efficiency losses due to radial variations may be tempered. In the extreme, the twist on the fan blades could be optimized for the radial variations of the flowfield. Obviously, the complex mixing processes would need to be well understood with both test data and computational fluid dynamics simulations to support such an effort. Nonetheless, such work may be necessary in order to make the BLI concept a reality. 78 6.5 Summary and Additional Thoughts Overall, the ramifications of distortion on the compression system will at minimum require an additional stage on the high-pressure compressor and a significant speed increase to the fan relative to a traditional installation. Perhaps a complete redesign of the fan and low-pressure spool may be necessary to maintain acceptable efficiency. The system effects of these changes may not be minimal, with implications on the engine size and weight as well as development cost. Furthermore, here only the steady state aspects of distortion are treated with no mention of the impact of takeoff and rotation. Takeoff traditionally represents the most severe condition for engine stability due to distortion stemming from high angles-of-attack and engine internal clearances being at undesirable levels [10]. With a BLI installation, this problem may be amplified thereby requiring additional measures to rectify. Therefore, the aggregate impact of the distortion will have to be judged according to additional metrics. Nonetheless, one can see that the compression system for a BLI ingesting aircraft will be considerably different than that for a traditional installation. 79 7. Additional Considerations 7.1 Mechanical Design The presence of distortion for the BLI propulsion system has ramifications on the mechanical design of the static structure of the engine. The same physics that leads to instability of the compression system also serves to load the compressor blades in an asymmetrical manner. This unbalanced loading incites cyclic fatigue of the blading as the structure is strained and relaxed as it passes through regions of high and low stagnation pressure [10]. While for the stability argument the conjecture is that the distortion is communicated throughout the entire compression system, here the cyclic fatigue is mostly an issue for the fan. This is because owing to the large span of the fan blades the cyclic induced bending loads are more severe. As evidence, the bending stress for a rotating blade can be written as: (Eq. 7-1) Here s represents the solidity, t the blade thickness, and RT the tip radius. Given the larger tip diameter of the fan relative to the compressor, the blade root loading will be higher for the fan [9]. When considering the design ramifications it is clear that a first order assumption is the fan blades need to have increased mass to absorb the higher strains. More massive blades have thicker roots and perhaps even additional blades (higher solidity) are required to reduce the stresses to acceptable levels. A higher fan mass influences the size of the fan hub that holds the blades. The higher centrifugal stresses tend to require a more massive fan hub accordingly. As a consequence, the shaft that drives the entire assembly needs to be enlarged to handle the increased inertial loads. Also the bearings and their locations will to be changed. Overall the entire low spool becomes heavier to provide the required robustness and in the process a significant rotor dynamics problem is created. A complete redesign of the low-pressure spool mechanical system is a real possibility. 80 Alternatively, doing nothing, results in the fan blades being subjected to high levels of cyclic stress leading to instances of high cycle fatigue (HCF) resulting in failure of the fan rotor assembly. This represents a serious flight safety issue. At a minimum, the HCF problem requires considerable maintenance costs to monitor and replace parts as necessary. To obtain a feel for the possible implications of distortion on the low-spool weight increase consider the pie chart in Figure 7.1 Remaining Engine Components Fan 37% 33% Remaining Low Spool Components 30% Figure 7.1: Turbofan Weight Summary [11] Here one sees that the low spool contributes 63% to the total weight of the engine. Of that 63%, the fan weight contributes about 50%. Consequently, any change to the fan and or low spool will have a significant impact on the total weight of the propulsion system and therefore on the fuel burn of the aircraft for a particular mission. 81 7.2 Engine Noise In today's aviation environment noise is a preeminent design concern. With ever increasing air traffic and the encroachment of airports into residential areas the noise impact of aviation is felt on a greater percentage of the populous. Noise is a significant nuisance and can limit the operations of aircraft thereby affecting the economic potential for the operator. Quieter aircraft will have a fundamental advantage as noise regulations continue to become more stringent in the future [14]. In general the noise produced by a turbofan engine can be classified into two major groups: 1) Exhaust jet noise and 2) Turbomachinery noise. Figure 7.2 illustrates this point: Fan Noise Fan Exhaust Noise Core Exhaust Noise Figure 7.2: Sources of Engine Noise The distinguishing feature of a highly embedded BLI propulsion system would be in the level of fan noise projected out the front of the engine. This as the result of the longer ducts required to feed the engine and the "S" type bends that are necessary as the engine 82 system provides greater ability to make use of Helmholtz resonators that serve as acoustic dampers. The additional surface area of the ducting allows more space for such devices. Also, a positive coupling exists between the embedded propulsion system and the airframe noise. Here the trailing edge noise is reduced in proportion the boundary layer air captured. Furthermore, less interference noise is created without the pod-pylon interaction. In all, the BLI embedded propulsion system has the potential to be quieter than the pod/pylon installation [10]. 7.3 Cost Implications When considering the cost implications for the highly embedded BLI propulsion system one must distinguish between two main types of cost: 1) Engine acquisition cost and 2) Operations cost of the in-service propulsion system. Engine acquisition cost is the cost to the airframer for the purchase of the engines and is on the order of 5 - 10% of the cost of the aircraft. This cost is representative of the development, manufacturing, and certification and testing resources expended by the engine maker. Operations cost encompasses the engine's fuel consumption and maintenance related expenditures while in revenue service. Together, I and 2 combined represent about 20% of the total operating cost for an aircraft. Figure 7.3 illustrates the cost breakdown for a traditional revenue service airliner. The influences of the BLI propulsion system on the two aspects of cost are treated in turn. Here only a qualitative investigation of cost is attempted with the goal of highlighting some of the salient aspects which will factor into the cost differences from a traditional propulsion system installation. 83 Fuel Engine 11% Maintenance 2% Engine Indirect Ownership Operating Costs 6% 42% Flight Crew 8% Maintenance 4% Airframe Ownership 27% Figure 7.3: Cost Breakdown [11] 7.3.1 Engine Acquisition Cost The development cost of the engine for a BLI configuration is presumably larger than that for a traditional installation. This conclusion stems from the higher technology levels implemented and the overall unprecedented nature of the concept. The technical challenges to such an engine installation are numerous with the most striking of these being the very high level of inlet distortion. As has been discussed, the first order defense against distortion related issues is to build in additional margin into the compression system design. However, design margin alone does not treat the additional problems associated with power transients, aircraft maneuvers, and other destabilizing effects. These issues need significant development work in order to make such a demanding compression system operationally viable from both a performance and safety perspective. In addition, the efforts to combat the high cycle fatigue problems stemming from the distortion require new technology development programs with associated research and 84 testing. In total, the compression system design represents a significant departure from the standard commercial application. Consequently, this is reflected as higher development cost. The unprecedented nature of the propulsion system impacts the testing and certification for the candidate engine concept. Commercial engine testing for certification is a rigorous process where the safety of the engine is proven under a variety of extreme operating conditions. For a BLI configuration new testing procedures need to be developed commensurate with the different operating conditions of the engine and aircraft. These changes are instilled at a cost to the engine manufacturer and may not be trivial. Certification for an engine takes years and cost hundreds of millions of dollars. Therefore, large perturbations or additions to the process have a very drastic impact to the delivery cost of the engine. Commercial engines are typically priced on a per pound of thrust basis. The higher technology requirements and new testing procedures will tend to increase the cost per pound of thrust for a BLI engine in comparison to a traditional engine. Figure 7.4 illustrates the type of shift which could be expected: 85 170- Higher Technogy & New Certification Procedures drives cost up 130 - 0 ; 110- 70- 50 0 20 40 60 80 100 120 140 Thrust Class -lbs/1000 Figure 7.4: Engine Cost per Pound of Thrust [11] 7.3.2 Engine Operations Cost The operations cost represents the fuel consumption and maintenance requirements for the propulsion system. The cycle optimization indicated a 4.2% reduction in the realizable fuel burn for the BLI concept. In lieu of any other mitigating factors the fuel burn reduction indicates a major cost savings to the operator of the aircraft. However, the highly embedded propulsion system has the potential for higher maintenance costs owing to: 1. Reduced accessibility resulting from embedding the engines in the fuselage. 2. Increased maintenance stemming from distortion-induced problems. 86 Accessibility is a critical issue for maintenance on commercial aircraft. The ability to rapidly perform routine maintenance allows the aircraft to remain on schedule and therefore maximize the amount of revenue service and generate airline profit. As a general rule, the more integrated the engine is with the aircraft the more difficult it is for maintenance personnel to perform their jobs [10]. The BLI propulsion system, with its high degree of integration, is more difficult to access. This results in longer maintenance durations and hence maintenance man-hours. With good systems integration the accessibility issue is minimized but presumably it is not as favorable as the traditional configuration with a pod and pylon installation. Given the predisposition to high cycle fatigue, BLI engines need to be serviced more often to ensure that no flight safety risks are present (i.e. cracking in the turbomachinery). In the extreme, the engines may have less time on wing owing to the risks inherent to HCF. Increasing the required maintenance directly drives up maintenance man-hours and therefore cost. Reducing the engine time on wing affects the revenue production capability and therefore represents lost profit. This indirect cost proves most important if the amount of servicing required is significantly increased. Overall, the complete cost picture needs a deeper investigation in order to fully understand the ramifications. While the fuel burn benefit is clear the more nebulous maintenance and engine development costs need additional investigation. The answer to the question of cost is essential to quantifying the benefits of the concept. In the end it is cost that represents the distinguishing characteristic that determines whether the concept is a success or failure. 87 8. Conclusions and Future Work 8.1 Summary The design space for the highly embedded propulsion system utilizing BLI is framed with respect to a traditional pod and pylon installation in terms of fundamental engine performance and design metrics. To this end, the trends between these two propulsion system configurations are identified and quantified for the particular instance of the BWB commercial transport. From the parametric cycle analysis, the need for larger propulsion systems is evident. As a consequence of the reduced pressure recovery, the specific thrust of the engine cycle is lower. Therefore, the fan diameter of the engine increases owing to the higher airflow requirement of the engine. In turn, the gas generator core size is larger to provide the necessary horsepower to drive the larger propulsor. The result is a propulsion system with a reduced thrust-to-weight ratio. In addition, the overall efficiency of the propulsion system is reduced which is reflected in the higher specific fuel consumption of the engine. Overall, the performance of the embedded engine is reduced with respect to the traditional modular installation. Using a simple boundary layer model, a study is conducted to determine the trend of fuel burn reduction due to the torpedo effect with propulsive cycle selection (bypass ratio). Here the inherent weight reduction of embedding, owing to pylon removal, is invested into the propulsion system to increase the fan diameter and the bypass ratio. From this analysis, the optimum engine size trend is towards larger fans (airflows) to capture the increasing benefit of BLI, until the point where the additional engine weight outpaces the profile drag reduction benefit. The embedded configuration, with the higher bypass ratio, has higher propulsive efficiency which augments the fuel burn benefit stemming from the drag reduction. By allowing the fan diameter to increase from that used for the traditional pylon/pod configuration, an additional 1% fuel bum reduction is realized. Overall, the study predicted a maximum 4.2% reduction in fuel burn when the entire embedded weight savings is put back into the engine. 88 The distortion impact on the compression system design is evaluated using empirical data and a first order rationalization of the distorted flowfield. The loss of surge margin is calculated owing to the total pressure distortion resulting from the incomplete boundary layer mixing. From this, one additional stage to the high-pressure compressor is required in order to maintain the necessary surge margin for safety and operability. The fan speed is increased about 4% to provide adequate margin in keeping with a single stage design. In addition, the mechanical design ramifications of distortion are investigated. The implication here is the need for a heavier, more robust low-pressure rotor to absorb the vibration induced loadings. With the low spool comprising about 60% of the weight of the total engine, any weight increase will be significant. Finally, a look into the cost ramifications for the embedded engines is conducted. Here cost is divided into the engine acquisition cost and the engine operations cost. Engine acquisition cost is higher owing to the greater development and testing cost for the novel, unprecedented concept. Engine operations cost is higher or lower depending on the maintenance impact of the embedded engines. With a 4.2% reduced fuel bum, operations cost is lower. However, if the engines require increased maintenance and/or are less accessible, the maintenance cost could offset any gains from fuel burn. Overall, more insight is needed into the maintenance aspect in order to more fully answer the question of cost. The analysis of the engine subsystem has determined a set of salient aspects which will be important when considering any boundary layer ingesting aircraft concept. While the engines will not be a mitigating factor in such a concept, considerable care will need to be provided so as to adequately handle the integration issues. What is clear is that engines designed for a traditional pod/pylon installation will not be the best choice for a BLI configuration. New engine designs will need to be developed that will more optimally fit the performance constraints and the design space. 89 8.2 Recommendations for Future Work This project focused solely on the engines and the considerations for the propulsion system design. The question that remains to be answered is does a BWB with a novel BLI propulsion system make sense from an overall systems perspective. To answer this question the impact on the airframe performance and the system cost needs to be determined For the airframe analysis the influence of BLI on the aircraft aerodynamics is a primary interest. This includes determining the wetted area (profile drag) and trim drag reductions from embedding. Also, the impact of the engine flowfield on the span loading should be investigated to quantify any lift-induced drag changes. In addition, exploration of functional integration benefits stemming from embedding should be pursued. This includes expanding the use of existing aircraft structure in the rear of the aircraft to more efficiently provide for airframe-engine integration. For a commercial aircraft to be successful cost must be minimized. Therefore in order to determine the system benefits of BLI the impact on total system cost is essential. To this end more work is needed to understand the maintenance cost implications for BLI propulsion systems. This would include both the accessibility issue and the possibility of more frequent maintenance intervals. In addition, the manufacturing and assembly benefits of the highly integrated configuration should be reflected in terms in cost figures of merit. 90 References 1. Rodriguez, David L., "A Multidisciplinary Optimization Method for Designing Boundary Layer Ingesting Inlets", Doctor of Philosophy Thesis, Stanford University, 2001. 2. Murthy, S.N.B., Paynter, G.C., "Numerical Methods for Engine-Airframe Integration", Progress in Astronautics and Aeronautics, Volume 102, 1986. 3. Hill, P., and Peterson, C., "Mechanics and Thermodynamics of Propulsion", Addison- Wesley Publishing Co., 1970. 4. Mattingly, K., "Elements of Gas Turbine Propulsion", McGraw-Hill Publishing Inc., 1996. 5. Lissaman, P., "Boundary Layer Induction Effects and Drag-Thrust Bookkeeping", BWB Research Note USC 3, 1995. 6. Smith, L. H., "Wake Ingestion Propulsion Benefit", Journal of Propulsion and Power- Volume 9, Jan.-Feb. 1993. 7. Smith, A.M.O., and Roberts, H.E., "The Jet Airplane Utilizing Boundary Layer Air for Propulsion", Journal of Aeronautical Sciences-Volume 14, Feb. 1947. 8. Douglass, W.M., "Propulsive Efficiency with Boundary Layer Ingestion", McDonnell Douglas Report MDC J0860, 1970. 9. Kerrebrock, J. L., "Aircraft Engines and Gas Turbines", The MIT Press, 1992. 10. Oates, Gordon. C., "The Aerothermodynamics of Aircraft Gas Turbine Engines", The Air Force Aero Propulsion Laboratory, AFAPL-TR-78-52. 11. Pratt & Whitney Internal Presentation 12. Longley, J. P., and Greitzer, E.M., "Inlet Distortion Effects in Aircraft Propulsion System Integration", XXXX. 13. Koch, C.C., "Stalling Pressure Rise Capability of Axial Flow Compressor Stages", Journal of Engineering for Power - Volume 103, 1981. 14. Making Future Commercial Aircraft Quieter, NASA Facts, Lewis Research Center FS-1997-07-003-LeRC. 15. Liebeck, R.H., "Design of the Blended-Wing-Body Subsonic Transport", ALAA No. 2002-0002. 16. http://adg.stanford.edu/aa241/structures/weights.html 17. Reid, C., "The Response of Axial Flow Compressors to Intake Flow Distortion", Gas Turbine Conference, Cleveland, OH, 1969. 18. Greitzer, E., "The Stability of Pumping Systems", Journal of Fluids Engineering, Vol. 103, June 1981. 91 Appendix 1: Project Timeline Overall Design Project Timeline Assess Literature 6 Generate Concepts 5 Select Preferied CDnfiguration 4 Decompcise Project 3 PerformAnalysis 2 J Merge 1 I I I I I I I |0 1 1 | 1 | 1 |2| 1 [ 1| I I I 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 5 March 10 Aug Time (weeks) 92 Appendix 2: PropulsionSystem DesignTimeline PropulsionSystem Design Timeline Engine model acquisition & setup 7 Excel tool setup 6 Parametric cycle anal) ,sis 5 Imbedded engi ie trade study 4 Downs6lect optimum ycle 3 Investigateoperability impact 2 E investighte noise& costimplications 1 -J * ______I I I 1 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 27 May 1 Oct Time (weeks) 93 Appendix 3: Boeing Project Letter 10 May 2002 Dr. Robert Liebeck The Boeing Company 2401 E. Wardlow Rd. MC C078-0316 Long Beach, CA 90807-5309 Dear Dr. Liebeck: During our 20 March MEng Project Design Review we characterized our BWB Highly Integrated Propulsion System Study success goal as having two primary elements. The first element involves the determination of a preferred embedded propulsion concept and the second deals with quantifying the performance and other trade issues related to that concept compared to the pylon-pod configuration baseline. We have generated several integrated propulsion concept variants and placed them in a Pugh Matrix. The matrix symbols indicate how each embedded concept compares to the pylon-pod basline. A plus indicates "better than," a minus is "worse than," and an "S" means it is the same. By summing the symbols we can show our logic for an initial selection of a preferred embedded concept It's important to note that no numerical weighting scheme has been used here and we think this is consistent with the early stage of our project. Instead we are using these abstract symbols to make an argument for which embedded concept becomes the basis for our quantification effort. The matrix represents our best effort at characterizing the anticipated performance and other trade issue trends. We are sending the matrix and concept drawings to you with the hope that you will circulate this package to key engineers on your BWB Project. We ask that they comment and mark-up the package. By including expert opinions from engineers who have been very close to BWB-type issues - we hope to improve the chances for selecting the best preferred embedded concept. Please return the package to us one-week after you receive it and we will move out on the second quantification phase of our project. Thank you in advance for helping us. Chris Hanlon & Vivian Shao Room 33-409 (ical Fran Marrone) Massachusetts Institute of Technology 77 Massachusetts Avenue Cambridge, MA 02139 94